Articles by tag: think

Articles by tag: think

    Finishing the Chassis

    Finishing the Chassis By Kenna and Janavi

    Task: Build a Chassis

    We have been working on this chassis for over 3+. In out last post, we had thought the wheels were ready to go. However, various parts, such as wheel mounts, had been put on backwards or were unusable so we had to do everything over again.
    Now that the robot has wheels, we started on attaching the REV expansion hub and battery. The chassis is square, but has an asymmetrical structure of tetrix bars. Attaching the battery was the simple part since previous version of the robot had a 3D-printed battery holder that would be screwed on. There was no way to effectively place the expansion hub on the tetrix rails. Instead, we attached a thin plank of wood to two parallel bars, drilled a couple holes, and screwed the hub on.
    Overall, it is a very no-frills chassis. We had to cut most of the side shields off because they were becoming more of an obstruction than an aid.

    Next Steps

    Though the physical robot has been built, it has no code. Both of us will be learning how to program a basic pushbot.

    UIL 2018

    UIL 2018 By Abhi, Karina, Evan, Janavi, Austin, Justin, and Shaggy

    Task: Attend the 2018 UIL Robotics Competition

    Background

    For those who don't know, UIL Robotics is the premier state robotics competition for Texas. Iron Reign has been a beta-testing partner since its inception, and this year was the event's first year as a full-fledged program.

    To participate in UIL, a team must win at a Regional level, and have a good overall showing. This year, since we got 2nd Inspire at Regionals and 3rd Inspire at Oklahoma Regionals, we were a shoo-in for an invitation. Being a state event, the DISD STEM Dept. supported us through transportation, food, and lodging along with other DISD teams such as Mechanicats.

    The Night Before

    As with all Iron Reign tournaments, we stayed up way longer than we should have. But, unlike other times, we had a purpose: to help fellow teams.

    We assisted the other DISD team, Mechanicats with programming and driver practice. In particular, they didn't have a working autonomous to begin with. But, with our half-field and glut of programmers, we helped them create a basic autonomous for the next day. As well, we collaborated on their TeleOp to make it more driver-friendly.

    The Day Of

    We walked into the tournament, tired, but excited for the last tournament of the season, led by our two robots, Kraken and C.A.R.T. BOT. Kraken is our Relic Recovery robot; a tank on wheels with specially cut aluminum sideplates and our proprietary REVolution system. So, it got plenty of looks. Then, we also brought the newest addition to the Iron Reign family: CART BOT. CART BOT is the automated corpse of our robot cart. For the past month, we've been tearing it down, replacing its wheels, motorizing it, adding a power source, and so much more. It tops out at 20 MPH and can carry 300 lbs. without blinking an eye. Naturally, we thought UIL was the perfect place to bring it out.

    Since UIL is the last tournament of the season and has no real consequences, we use it as a trial field for next year's changes. First, we had Evan lead our pit crew team as practice for next year. As well, we used the competition to practice driving for next year as well as improve our scouting strategies after worlds.

    One of the best things about UIL is the ability to really interact with other Texas-area teams that we normally wouldn't see until Supers. A lot of the teams came over to see our robot, which is kind of understandable because it's probably the best robot we'll ever build. But, we had a surprising number of teams come up to talk to us about our Engineering Journal, including people who had already seen our journal online and wanted to talk about it to us in person (Vitruvian Voltage).

    Robot Performance

    Even though we enjoy UIL, it's never our best competition of the year. Some of this is due to exhaustion; we tend to run out of steam by then, but it can also be attributed to that UIL is a robot-game intensive event, and Iron Reign tends to focus more on awards. So, we tend to comparatively underperform as compared to a theoretical Iron Reign stand in.

    We started off the day in a bad place, as one of the chains on the robot snapped for the first time in the season. However, we still managed to win the match as we were carried by our partner. But, we managed to do decently in the next four matches. This wasn't entirely due to luck, it was just that we had more competition experience than some of the other teams due to Worlds, and were able to perform more effectively.

    Luckily, our scouting paid off, and we were chosen as the first pick of the #1 alliance. We won our first final match, but then lost the next two due to unreliability.

    The UIL Difference

    Unlike FTC, UIL puts much less of an emphasis on judging. First, there aren't any presentations: everything is done at the pit. In addition, UIL judges are FRC first, and FTC second, so they weren't aware of many differences between the two. Finally, the awards mean nothing.

    Next Steps

    This was the last competition of the season, so now Iron Reign will go into Funding, Outreach, and Recruitment mode for a while for the next season, but keep track of our blog to see what we'll do next. Relic Recovery '17-'18, signing off.

    Swerve Drive Experiment

    Swerve Drive Experiment By Abhi

    Task: Consider a Swerve Drive base

    Last season, we saw many robots that utilized a swerve drive rather than the mecanum drive for omnidirectional movement. To further expand Iron Reign's repertoire of drive bases, I wanted to further investigate this chassis. Swerve was considered as an alternative to swerve because of its increased speed in addition to the maneuverability of the drive base to allow for quick scoring due to its use of traction wheels at pivot angles. Before we could consider making a prototype, we investigated several other examples.

    Among the examples considered was the PRINT swerve for FTC by team 9773. After reading their detailed assembly instructions, I moved away from their design for many reasons. First, the final cost of the drive train was very expensive; we did not have a very high budget despite help from our sponsors. If this drive train was not functional or if the chassis didn't make sense to use in Rover Ruckus, we would have almost no money for an alternate drive train. Also, they parts used by 9773 involved X-rail rather than extrusion rail from REV. This would cause problems in the future as we would need to redesign the REVolution system for X-rail.

    Another example was from team 9048 which appeared to be more feasible. Because they used REV rail and many 3D printed parts, this was a more feasible prototype. Because they didn't have a parts list, we had the find the rough estimate of cost from the REV and Andymark websites. Upon further analysis, we realized that the cost, though cheaper than the chassis of 9773, would still be a considerable chunk of our budget.

    At this point it was evident most swerve drives being used are very expensive. Wary of making this investment, I worked with our sister team 3734 to create a budget swerve with materials around the house. A basic sketch is listed below.

    Next Steps

    Scavenge for parts in the house and Robodojo to make swerve modules.

    Swerve Drive Prototype

    Swerve Drive Prototype By Abhi and Christian

    Task: Build a Swerve Drive base

    Over the past week, I worked with Christian and another member of Imperial to prototype a drive train. Due to the limited resources. we decided to use Tetrix parts since we had an abundance of those. We decided to make the swerve such that a servo would turn a swerve module and the motors would be attached directly to the wheels.

    Immediately we noticed it was very feeble. The servos were working very hard to turn the heavy module and the motors had trouble staying aligned. Also, programming the chassis was also a challenge. After experimenting further, the base even broke. This was a moment of realization. Not only was swerve expensive and complicated, we also would need to replace a module really quickly at competition which needed more resources and an immaculate design. With all these considerations, I ultimately decided that swerve wasn't worth it to use as a drive chassis at this time.

    Next Steps

    Consider and prototype other chassis designs until Rover Ruckus begins.

    Big Wheel Ideas

    Big Wheel Ideas By Janavi

    Task: Create a Unique Chassis

    This summer, we're working on creating unique chassis that are outside of our comfort zone. Often we choose safe bases - opting for ones that we have tried in the past and know work. But, taking the opportunity to explore unique bases allows us to see their performance. One of our ideas is for a two-wheeled robot, with two large wheels and one, smaller, non-motorized omniwheel. We think that this 2-wheeled robot would be a good opportunity for Iron Reign, as we know that our robot has to be lighter than the Relic Recovery robot and a non-mecanum drive would be much lighter. Here is a drawing of what we plan the chassis to look like:

    To make this chassis the most efficient based on what we currently know about the competition (light weight robot needed) we are planning to do different tests and calculations to determine the proper motor-gear ratio needed and the wheel locations to properly balance the robot. We also need to perform tests to determine the best material to use for the robot. In the past we've used REV rails for the majority of our structure but due to the weight limit on our robot we plan to minimize metal in our design rather opting for materials that are just as functional but weight less.

    Next Steps

    Perform calculations comparing different motors as well as different wheel ratios to determine the optimal ratios

    Position Tracking

    Position Tracking By Abhi

    Task: Design a way to track the robot's location

    During Relic Recovery season, we had many problems with our autonomous due to slippage in the mecanum wheels and our need to align to the balancing stone, both of which created high error in our encoder feedback. To address this recurring issue, we searched for an alternative way to identify our position on the field. Upon researching online and discussing with other teams, we discovered an alternative tracker sensor with unpowered omni wheels. This tracker may be used during Rover Ruckus or beyond depending on what our chassis will be.

    We designed the tracker by building a small right angular REV rail assembly. On this, we attached 2 omni wheels at 90 degrees to one another and added axle encoders. The omni wheels were not driven because we simply wanted them to glide along the floor and read the encoder values of the movements. This method of tracking is commonly referred to as "dead wheel tracking". Since the omnis will always be touching the ground, any movement will be sensed in them and prevents changes in readings due to defense or drive wheel slippage.

    To test the concept, we attached the apparatus to ARGOS. With some upgrades to the ARGOS code by using the IMU and omni wheels, we added some basic trigonometry to the code to accurately track the position. The omni setup was relatively accurate and may be used for future projects and robots.

    Next Steps

    Now that we have a prototype to track position without using too many resources, we need to test it on an actual FTC chassis. Depending on whether or not there is terrain in Rover Ruckus, the use of this system will change. Until then, we can still experiment with this and develop a useful multipurpose sensor.

    Chassis Flyer

    Chassis Flyer By Ethan

    Kraken

    This is Iron Reign’s world-championship robot from last season. The basic rundown is this:

    • Weight - 42 lbs
    • Size - 18x17.8x17.5 inches
    • Drive - Mecanum
    • Main parts kit - REV

    Iron Reign uses two design processes in conjunction with each other to create efficient and reliable parts: iterative, continual improvement and competitive design.

    An example of these design processes working in conjunction is the process of designing our cryptobox intake system. One person had the idea to build an arm-style grabber seen on many current competition robots. His design, however, included shorter arms for space’s sake and a more compact lift system than normal. The second person decided to build a unique conveyor-belt system which used friction to hold blocks in space and move them vertically. Through the competition, we determined that the prior design was more efficient and took up less space than the latter, so we settled on his design, adding in a linear slide for lifting at the end of the process. Then, Kaizen comes in. Through firsthand experience in scrimmages, we learned that the grabber system isn’t as reliable as we thought when first testing. So, we have designed a new grabber system that moves like the arms did previously, but also rotate with soft spikes attached to hold blocks with friction better without damaging them.

    As this soft-spike system ceased to perform to our expectations, we looked to other mechanisms to pick up and deliver blocks effectively. We created a new grabber that still used the rotating systems of the soft-spike, but instead, we used custom 3D printed “octopuckers” which had a much tighter grip on the glyphs. As well, inside the gripper, we created a custom “lift” made out of NinjaFlex so that the blocks could be moved up and down internally in the gripper, eliminating our need for stacking.

    Later, we further improved upon the grabber design, attaching it to a conveyor belt so that we could move glyphs all across our robot in order to score higher, using our REVolution system. This is the most ambitious use of our REVolution system yet, and we strongly encourage the reading judges to view it at the pits.

    BigWheel

    The main purpose of this robot is to see if larger wheels will give us an advantage in the competition. Right now, we’re guessing that the competition field will have debris, and we hope that the large wheels will perform better in this environment.

    • Size: ~18x18 in
    • Wheels - 8in large, regular omni wheels in front
    • Part System: Custom parts

    Garchomp

    For skill development we have newer builders replicating the chassis portion of our competition robot (Kraken). This one will not be weighed down by the additional upper structure of the competition robot and so should be a closer comparison in weight class to most of the other chassis designs under consideration here. Garchomp has a simplistic design and is nothing more than mechanums, rev rails, motors, sprockets, wires, and a rev hub. The large mechanums are held together using side plates from the 2017-18 competition season. These are geared up to neverest 40:1 motors.

    • Size: ~18x18 in
    • Wheels: Mechanum
    • Part System: REV
    • Motors: Neverest 40:1

    Summer Chassis Project - July Meeting

    Summer Chassis Project - July Meeting By Kenna, Ethan, Charlotte, Karina, Shaggy, and Abhi

    Task: Compare & Collaborate on Chassis

    At Big Thought's offices in downtown Dallas, three teams met. Technicbots (Team 8565), EFFoRT (Team 8114), Schim Robotics (12900), and Iron Reign are all part of the North Texas Chassis Project. The goal is for each team to create any number of chassis and improve their building skills by learning from the other teams.

    The meeting began with an overview of all teams' progress. Each team presented their thought process and execution when creating each bot and discussed why/how everything was done. At the end, we all reviewed the rule changes for the 2018-19 season. Once all questions had been asked and answered, testing began.

    Austin Lui of Technicbots gets their chassis ready for testing.

    Using leftover tiles from last season, we set up a small field in Big Thought's blue room. Technicbots provided a ramp to do enhanced testing with. All teams plan on testing:

    • Forward speed
    • 3 second turn
    • Up/Down ramp
    • Balancing stone
    • Weight-pulling
    • Straight line drift
    • 90/180° turn offset

    Connor Mihelic of EFFoRT adds some finishing touches.

    We know from Google Analytics that our website has about 200 visitors a month but we rarely meet the people who read and use our blog posts. Today, we got to meet the mentors of Team 12900 from a middle school in Plano, TX. When they and their students were starting out as a team, they utilized our tutorials and journal. Apparently their teams members are avid followers of our team, which was very meaningful to hear. Some non-FTC friends visited as well and were introduced to cartbot.


    Terri and Grant Richards of Schim Robotics.

    Next Steps

    Using what we learned from the other teams, we will begin to improve all of our chassis. Most of them are at varying levels of completion so now we want to concentrate on getting all of them to the same level of functionality. Garchomp is, notably, the most behind so he will be getting the most attention from here on out.

    Replay Autonomous

    Replay Autonomous By Arjun

    Task: Design a program to record and replay a driver run

    One of the difficulties in writing an autonomous program is the long development cycle. We have to unplug the robot controller, plug it into a computer, make a few changes to the code, recompile and download the code, and then retest our program. All this must be done over and over again, until the autonomous is perfected. Each autonomous takes ~4 hours to write and tune. Over the entire season, we spend over 40 hours working on autonomous programs.

    One possible solution for this is to record a driver running through the autonomous, and then replay it. I used this solution on my previous robotics team. Since we had no access to a field, we had to write our entire autonomous at a competition. After some brainstorming, we decided to write a program to record our driver as he ran through our autonomous routine and then execute it during a match. It worked very well, and got us a few extra points each match.

    Using this program, writing an autonomous program is reduced to a matter of minutes. We just need to run through our autonomous routine a few times until we're happy with it, and then take the data from the console and paste it into our program. Then we recompile the program and run it.

    There are two parts to our replay program. One part (a Tele-op Opmode) records the driver's motions and outputs it into the Android console. The next part (an Autonomous Opmode) reads in that data, and turns it into a working autonomous program.

    Next Steps

    Our current replay program requires one recompilation. While it is very quick, one possible next step is to save the autonomous data straight into the phone's internal memory, so that we do not have to recompile the program. This could further reduce the time required to create an autonomous.

    One more next step could be a way to easily edit the autonomous. The output data is just a big list of numbers, and it is very difficult to edit it. If we need to tune the autonomous due to wear and tear on the robot, it is difficult to do so without rerecording. If we can figure out a mechanism for editing the generated autonomous, we can further reduce the time we spend creating autonomous programs.

    C.A.R.T. Bot Summer Project

    C.A.R.T. Bot Summer Project By Evan, Abhi, and Janavi

    Task: Enhance our robot-building skills

    At Iron Reign, we hate to waste the summer since it’s a great time to get all the ridiculous builds out of the way. Thus, we created C.A.R.T. Bot (Carry All our Robotics Tools). Our constant companion these last few seasons has been our trusty Rubbermaid utility cart which has been beaten and abused, competition after competition, as it carried all our tools and robots. Because of all of this, we decided it was time to show the cart a little love, and in typical Iron Reign fashion, we went all out and turned it into a robot.

    Our first step was to switch out the back wheels on it to elf-sized bicycle wheels, allowing us to take on the mightiest of curbs and motorize it. To attach the wheels, a four foot or so cylinder of threaded steel was inserted in holes on either side of the cart. Two slots were cut out in the bottom for the wheels and they were eventually slid on, but not after 3D printed mounts for sprockets were attached to the wheels, enabling us to gear them in a one to one ratio with the sprocket attached to the motors, which consisted of two SIM motors commonly found on FRC robots.

    Before we used SIM motors, we attempted to power the cart using two Tetrix motors which were geared for speed but, due to load, barely moved at all. Besides a lack of power, they also tended to come out of alignment, causing a terrible noise and causing the cart to come to a stall. This was quickly scrapped. To mount the motors, we used two pieces of aluminum bars and bolted them to the motors, then screwed them to the floor of the cart, aligned with the wheels. We chained them together and got about powering the system. We got two 12-volt batteries and chained them in parallel so as to not overload the system, and hooked them up to a REV hub. Then, we ran them through a switch and breaker combination. We connected the motors to the rev hub and once we had it all powered up, we put some code on it and decided to take it for a spin.

    It worked surprisingly well, so we went back in and put the finishing touches on the base of Cart Bot, mainly attaching the top back on so we could put stuff on top of it, and cutting holes for switches and wires to run through, to make it as slick as possible. We added a power distribution station to assist with the charging and distribute current to any device we decided to charge on the cart. We will eventually hook this up to our new and improved battery box we plan on making in the few spare moments we’ll have this season, just a quick quality of life improvement to make future competitions go smoothly.

    Next Steps

    Our cart box isn’t done yet, as we intend to make a mount for a solar panel, which we will be able to charge the cart during the downtime in competitions (only if there’s a good window we can park it next to). The cart wasn’t just about having a cool new and improved cart that we don’t have to push (which it is), it also was a test of our engineering skills, taking things that never should have been and putting them together to make something that we will utilize every competition. We learned so much during this experience, I for one learned how to wire something with two batteries as not to destroy the system, as for everyone else, I can’t speak for all but I think we learned a very important lesson on the dangers of electricity, mainly from the height of the sparks from an accidental short that happened along the way. Despite this, the cart came out great and moves smoother than I ever could have hoped. The thing is a real blast and has provided a lot of fun for the whole team, because yes, it is ridable. We predict the speed it’s set at is only a fifth of its full potential speed, and since it already goes a tad on the fast end we don't intend to boost it anymore while there’s a rider on it. Overall, the project was a success, and I’m personally very proud of my work as I’m certain everyone else is too. Come to see it at our table, I really think it’s worth it.

    Adjusting Garchomp's Chains

    Adjusting Garchomp's Chains By Janavi and Kenna

    Task: Build the Chassis

    In our last post, we thought that we had finished Garchomp. However, as we came back to the next practice, we realized that Garchomp's chains were incorrectly linked.

    So, we started to diagnose the problem. We noticed that the old REV rails we were using had dents in them, which caused the motors to shift, therefore causing the chains to come off the gears.

    To amend this problem, we decided to replace the REV rails ensuring that the motors would not shift during movement. To accomplish this we:

    • First, we loosened all of the screws on the current bar, carefully slid it out, and replaced it with new bars
    • Then we fixing all of the chains and confirming that each of them were individually working
    • we re-attached all of the cables to the robot
    • Ran a stress-tester program and hung the robot on a hook to allow us to properly observe the wheels
    Due to our tests we discovered that our wheels were running at different speeds, meaning that our robot constantly moved in circles. After checking that the motors were working, we discovered that it was our encoder cables that were plugged in wrong. After that, Garchomp began to run smoothly.

    Next Steps

    We will run more stress tests on our robot and make sure that it is up to par with our past robots.

    My Summer at MIT

    My Summer at MIT By Abhi

    Task: Spend a Summer at MIT

    Hello all! You might have been wondering where I went the entire summer while Iron Reign was busily working on tasks. Well for those of you interested, I was invited to spend a month at MIT as part of the Beaverworks program. I worked in the Medlytics course and analyzed medical data using machine learning methods. This seems distant from the work we do in FTC but I learned some valuable skills we could potentially use this season. But before I discuss that, I want to talk about the work I did while I was away.

    Traditionally, machine learning and artificial intelligence were used for enrichment of the technology. We have been seeing development of search engines to learn our searching trends and craft new results or online shopping websites like Amazon learning our shopping to suggest new items to buy. With the help of machine learning, all this has become possible but there are potential healthcare applications to the same technology. The new algorithms and techniques being developed have shown potential to save lives in times where traditional approaches had failed. Even with basic implementations of artificial intelligence, we have seen instances where a doctors provided an improper diagnosis while a machine said otherwise. These scenarios have further inspired research for medical analytics, which has become the focus of my course at MIT. The Medlytics course was dedicated to learn more about these issues and tackle some real world problems.

    The work I was doing was very intensive. I applied the algorithms we were being taught to a number of situations. One week, I was analyzing physiological signals to determine the state of sleep. The next week, I was training models to detect breast cancer from mammograms. Within all this work, the underlying structure was just techniques that could be applied to a number of fields. That brought me to think about the potential applications of my work in FTC. The neural networks and similar models I was training learned a number of scenarios of images or signals. I realized that by integrating computer vision, I could come up with something similar in FTC.

    To demonstrate an example of where this could potentially leave an impact, I will go with object detection. Right now, Iron Reign captures a series of images of the object of interest (an example is a cryptobox from Relic Recovery) and attempts to manually fine tune the OpenCV parameters to fit the object as accurately as possible. This sort of task could easily be delegated to a Convolution Neural Network (CNN) architecture. What is a CNN you ask? Well here is a brief description.

    In essence, the model is able to determine a pattern in an image based on edges and details. The image is processed through a series of layers to determine the shapes in the image. Then the model attempts to label the image as seen above with the car. If this was brought into context of FTC, we could train model to learn the shapes of an object (for example a wiffle ball) and then feed the information to the robot. The bot could then navigate to the object and pick it up. There are a vast number of applications to this, with this just being one. I hope that my knowledge can be of use for Rover Ruckus.

    Next Steps

    Wait for Rover Ruckus reveal to see if I can combine my expertise with new code.

    BigWheel CAD

    BigWheel CAD By Ethan

    Task: Create a mockup for BigWheel

    We've been working on a design for the chassis workshop for quite a while now. We already presented it at the first meeting, and now we need to work on the other components of our presentation: the weight testing, torque calculations, speed testing, and finally, a chassis model. To do the last one, we made a simple model in PTC Creo.

    Bigwheel Presentation

    Bigwheel Presentation By Arjun and Karina

    Task: Present about Garchomp

    As a new freshman on Iron Reign, I took on the responsibility of a robot we called Bigwheel. Karina and I worked on getting the robot into something that could be put through load tests, meaning tightening the chain, fixing misaligned sprockets, and getting the wiring together. We participated in the Chassis Presentation workshop hosted by technicbots for teams all around the North Texas region to work on one or more chassis, perform various tests with them and then present their findings. We presented our chassis Bigwheel, which is driven by 2 large 8-inch wheels, with a pair of 2 free-spinning Omni wheels in the back. This can be seen in the presentation below:

    To create our chassis we used 2 8-inch wheels, each driven by 2 Neverrest 60 motors. There are also two free-spinning omni wheels in the back. The robot uses REV rails and plexiglass for it's main body.

    Our first test is the 5-second distance test. Our robot had a lot of torque due to the Neverrest 60 motors, so it moved slower than other robots, but was unaffected by the additional 30lbs weight.

    Our second test is the 3-second turn test. Again, some other robots could turn better faster than us. However, due to having no proper mechanism for restraining our weights, along with other mysterious problems such as battery disconnections that only happened during this test, we were unable to try this test with load, however we presume that due to the torque, the results should be similar to those without load. Our center of rotation is also off due to only the front two wheels being powered. As such, the back of the robot makes a wide arc as it turns.

    Our next few test results are unremarkable.

    Our robot had a lot of sideways drift, mostly due to bad build quality. If we intend to use it during the season, we will try to fix this.

    Overall, our chassis performed well under load, but could use a little speed boost. If we want to further develop it, we plan to use Neverrest 20s with more torque on our external gear ratio, so we can get more speed out of it.

    Garchomp Presentation

    Garchomp Presentation By Janavi and Kenna

    Task: Present in the Inviational Presentation Series

    Today, we participated in the Chassis Presentation workshop for teams all around the North Texas region; the project was to design robots and perform various tests with them, then present findings. We presented our chassis, Garchomp, a mechanum wheel chassis as can be seen in the slide photos below.

    Presentation

    To create our chassis we used 4 never rest 40 motors one for each wheel and the structure of the chassis was created by using tetrix rails. We connected the wheels to the motors by using a 1:1 gear ratio. While there are many benefits to using a gear ratio for your wheels be forewarned that if your wheels are not perfectly aligned attaching your chains to mechanum wheels will become a living nightmare as can be seen in our previous posts.

    One of the reasons that attaching the chains was so difficult for us was because we discovered that because we had used wooden sides instead of the aluminum sides that Kraken used our wheels became misaligned to the two different types of wood used for the sides.

    While our robot is not able to do a 360 degree turn as fast as some other robots presented today it is able to hold a considerable amount of speed while moving at a constant speed.

    Since this chassis was designed for last years competition it is able to consistently drive onto the balancing stone

    Post Kickoff Meeting

    Post Kickoff Meeting September 08, 2018 By Karina, Charlotte, Ethan, Evan, Kenna, and Abhi

    Meeting Log September 08, 2018

    Today Iron Reign attended the FTC 2018-2019 season kickoff at Williams High School. After the event, we gathered back at our coach's house to talk about how we might approach this season's challenge. We welcomed prospect team members as well. They joined us in reviewing the reveal video and the games manuals.

    Today's Meet Objectives

    We wanted to have an understanding of the game design so that we could start going over robot designs. To do this we:

    • Watched the reveal video
    • Skimmed through game manual 1 and the preview of game manual 2

    Until we receive the field elements, we will have to plan and strategize using the resources listed above.

    Because we also had new possible team members over, we set expectations for this year. Actively recording our progress and blogging for the engineering journal was heavily stressed. We recognize the importance of having a good engineering journal and how it can help us advance. Our coach's house, the place where we have our meetings, is also cleaner than it has been in a long time after an intense cleaning session. Having an organized space maximizes efficiency, especially with the a larger team. Therefore, we expect for all team members to clean up after themselves and maintain the organization.

    Before we could discuss robot build ideas, we talked strategy. Parking in the crater and the landing zones will undoubtedly be easy to do. Since we know that designing a way for our robot to be able to lift itself onto the lander will be a more interesting challenge and will score us the most points, we will prioritize working on prototypes mechanisms for this task. Finding a way to gently lower down form the lander may be difficult. We will have to consider ways to not damage the robot, wiring, etc. We also agreed that it would make the most sense to have one mechanism that latches onto the hook on the lander, grabs gold and silver elements from the crater, and places these elements into the columns.

    Other topics we talked about include drive trains, problems with trying to create a mechanism that grab both the silver balls and gold blocks, as well as how we would be able to grab them out of the crater without going over the edge of the crater and getting stuck.

    Also, in previous seasons, we have had strong autonomous game, and so we decided to make the tasks in autonomous another top priority. We had our coders start discussing the field path for autonomous. Unfortunately, we will not be able to launch our team marker into the team depot.

    After the end of last season, a storm passed through and turned over shelves, trashing the robo-dojo. Some of our team members cleaned up the tent this afternoon. While it may not seem very important at the moment, this will be very helpful later in the season once we get our field elements for this year's challenge and want to set the field up. While cleaning, they also uncovered old, rusted metal tools and and pieces, which we will now be able to repair and save for future use.

    Besides helping with cleaning the tent, the new members showed a lot of interest in the game as well. They were eager to start building, and actually started creating prototype mechanisms for picking up the silver and gold elements.

    Today's Work Log

    Team MembersTaskStart TimeDuration
    KarinaWorking on blog2:004 hrs
    AbhiAutonomous planning2:004 hrs
    EvanRobot brainstorming2:004 hrs
    CharlotteRobot brainstorming2:004 hrs
    EthanWorking on blog2:004 hrs
    KennaCleaning tent2:004 hrs

    Rover Ruckus Brainstorming & Initial Thoughts

    Rover Ruckus Brainstorming & Initial Thoughts By Ethan, Charlotte, Kenna, Evan, Abhi, Arjun, Karina, and Justin

    Task: Come up with ideas for the 2018-19 season

    So, today was the first meeting in the Rover Ruckus season! On top of that, we had our first round of new recruits (20!). So, it was an extremely hectic session, but we came up with a lot of new ideas.

    Building

    • A One-way Intake System

    • This suggestion uses a plastic flap to "trap" game elements inside it, similar to the lid of a soda cup. You can put marbles through the straw-hole, but you can't easily get them back out.
    • Crater Bracing
    • In the past, we've had center-of-balance issues with our robot. To counteract this, we plan to attach shaped braces to our robot such that it can hold on to the walls and not tip over.
    • Extendable Arm + Silicone Grip

    • This one is simple - a linear slide arm attached to a motor so that it can pick up game elements and rotate. We fear, however, that many teams will adopt this strategy, so we probably won't do it. One unique part of our design would be the silicone grips, so that the "claws" can firmly grasp the silver and gold.
    • Binder-ring Hanger

    • When we did Res-Q, we dropped our robot more times than we'd like to admit. To prevent that, we're designing an interlocking mechanism that the robot can use to hang. It'll have an indent and a corresponding recess that resists lateral force by nature of the indent, but can be opened easily.
    • Passive Intake
    • Inspired by a few FRC Stronghold intake systems, we designed a passive intake. Attached to a weak spring, it would have the ability to move over game elements before falling back down to capture them. The benefit of this design is that we wouldn't have to use an extra motor for intake, but we risk controlling more than two elements at the same time.
    • Mechanum
    • Mechanum is our Ol' Faithful. We've used it for the past three years, so we're loath to abandon it for this year. It's still a good idea for this year, but strafing isn't as important, and we may need to emphasize speed instead. Plus, we're not exactly sure how to get over the crater walls with Mechanum.
    • Tape Measure
    • In Res-Q, we used a tape-measure system to pull our robot up, and we believe that we could do the same again this year. One issue is that our tape measure system is ridiculously heavy (~5 lbs) and with the new weight limits, this may not be ideal.
    • Mining
    • We're currently thinking of a "mining mechanism" that can score two glyphs at a time extremely quickly in exchange for not being able to climb. It'll involve a conveyor belt and a set of linear slides such that the objects in the crater can automatically be transferred to either the low-scoring zone or the higher one.

    Journal

    This year, we may switch to weekly summaries instead of meeting logs so that our journal is more reasonable for judges to read. In particular, we were inspired by team Nonstandard Deviation, which has an amazing engineering journal that we recommend the readers to check out.

    Programming

    Luckily, this year seems to have a more-easily programmed autonomous. We're working on some autonomous diagrams that we'll release in the next couple weeks. Aside from that, we have such a developed code base that we don't really need to update it any further.

    Next Steps

    We're going to prototype these ideas in the coming weeks and develop our thoughts more thoroughly.

    Testing Intakes

    Testing Intakes By Ethan and Evan

    Task: Design a prototype intake system

    In our first practice, we brainstormed some intake and other robot ideas. To begin testing, we created a simple prototype of a one-way intake system. First, we attached two rubber bands to a length of wide PVC pipe. This worked pretty well, but the bands gave way a little too easily.

    For our next prototype, we attached a piece of cardboard with slits to a cup approximately the size of a cube or block. It operates similarly to a soda cup lid with a straw hole. An object can go in, but the corners of the hole spring back so that it can't escape.

    Next Steps

    We probably won't go with this design - we'd have issues separating the different kinds of game elements, and it may be too slow to feasibly use. But, its a first step and we'll see what happens.

    Rover Ruckus Strategy

    Rover Ruckus Strategy By Ethan, Kenna, Charlotte, Evan, Abhi, Justin, Karina, and Aaron

    Task: Determine the best Rover Ruckus strategies

    Challenge Game Timing Points Level of Difficulty (1 - 3 [hard]) Priority Idea
    Landing Autonomous 30 2 Medium Latch attached to linear slides that allows us to descend rapidly
    Claiming Autonomous 15 1 High Autonomous, easy as bumping into wall
    Parking Autonomous 10 1 High Autonomous, just need to move
    Sampling Autonomous 25 2 Medium Autonomous, OpenCV solution as in similar years
    Latching End Game 50 3 High 3D-printed latch attached to linear slide strong enough to lift robot
    Robot in Crater End Game 15/25 1 High Driving
    Mining [Depot] Tele-Op 2 per item 1 High Rolling intake into box, then conveyor belt into the depot
    Mining [Cargo] Tele-Op 5 per item 2 High Long linear-slide arm that reaches the two feet into the lander with an intake/deposit on the end

    Choosing Drive Train

    Choosing Drive Train By Janavi

    Task: Analyze the game

    In our last post, we created a chart where we listed each task asked based on point value and the level of difficulty, separated by autonomous and teleop. Our goal is to find a drive train that will allow us to build a robot to accomplish all of these tasks efficiently and consistently, but this matrix will allow us to determine what to focus on first.

    Drivetrain Comparison

    This summer we created a variety of drivetrains for a summer chassis project hosted in coordination with other teams from the North Texas region. We have compiled a list of the drivetrains and the criteria we need to consider for Rover Ruckus.

    What do we need to look at in a Drivetrain?

    • Light
    • Sturdy
    • Easily Maneuverable
    • Fast
    • Low center of mass to avoid tipping
    • Reliability

    Comparison

    Eliminated? Reason for Elimination Pros Cons
    Miniature Mechanum Drive NO N/A
    • Omni-Directional
    • Fast turning
    • Easy to design
    • Experience with
    • Driving/Building
    • light
    Uneven power
    Big Wheel NO N/A Unique Design We have less experience
    Larger Mechanum Drive YES Need light robot; may use mini mechanum chassis instead Familiar Design Too heavy for this years competition
    Swerve YES Difficult design, Many motors and servos, we have less experience Easier to maintain at high speed Unfamiliar and difficult to design and maintain
    8-wheel Drive YES Many wheels, Difficult of maneuver, no omni directional movement 100% power forward Difficult to maneuver
    Holonomic Drive YES Less push power in all directions; hard to integrate into robot Easy to turn and maneuver Hard to design; hard to integrate into base; Only 50% power in all directions

    Selecting Wheels

    Selecting Wheels By Janavi

    Objective: Determine the type of wheel that best suits the chassis

    In the Choosing Drive Train E-16 we decided that we will use the chassis BigWheel. We know that our wheels need to be light weight but we need to determine the size of the wheel that will keep our robot far away enough from the ground for us to provide enough clearance to allow us to climb over the crater rim. But, if we choose wheels with a large radius we risk raising the center of mass.

    Pros Cons
    Ironton 12in. Solid Rubber Spoked Poly Wheel
    • light
    • durable
    • Large Turns
    • Extremely Large
    Ironton 16in. Solid Rubber Spoked Poly Wheel
    • light
    • durable
    • Raise center of mass
    • Extremely Large
    • To prevent tipping we now have a much shorter distance to correct imbalance
    Ironton 8in. Solid Rubber Spoked Poly Wheel
    • light
    • durable
    • Not large enough to significantly move the center of mass

    Brainstorming Two

    Brainstorming Two By Evan, Abhi, and Janavi

    Task: Have a 2nd brainstorming session

    We had another brainstorming session today, which allowed us to break down into some new building tasks.

    Intake System 3 - TSA Bag Scanner

    This part of our robot is inspired by the bag-scanning machine in TSA lines, more specifically the part at the end with the spinning tubes. The basic design would be like a section of that track that flips over the top of the robot into the crater to intake field elements.

    Intake System 4 - Big Clamp

    This one is self-explanatory. Its a clamp, that when forced over a block or a cube, picks it up. It's not that accurate, but it's a good practice idea.

    Lift 2 - Thruster

    We want to make lifting our robot easy, and we're thinking of a slightly different way to do it. For our new lift idea, we're installing a vertical linear slide that forces the robot upwards so that we can reach the lander.

    Next Steps

    We're working on building these prototypes, and will create blog posts in the future detailing them.

    Meeting Log

    Meeting Log September 15, 2018 By Charlotte, Karina, Kenna, Janavi, Evan, Abhi, Justin, and Ethan

    Meeting Log September 15, 2018

    Today Austin, an Iron Reign alumni, visited us from A&M! :)

    Today's Meet Objectives

    As our brainstorming and discussion continues, we are putting our ideas into action and making various prototypes and designs. We will continue to work with our new recruits and let them participate in a meaningful way with our building and in getting ready for competition.

    Today's Meet Log

    • Further brainstorming and discussion
    • Taking some inspiration from 30 hr robot reveal videos, we have continued the brainstorming for this year's robot. Our main subjects of discussion are our intake and lift, and some ideas that were thrown around were a conveyor belt-like intake and a lift that utilizes a linear slide which lifts the robot chassis. The details of our brainstorming session can be found at (E-19, Brainstorming Two - Enter the Void).
    • Prototyping and linear slides
    • Today, Abhi worked on a hook for hanging off the rover at first with Styrofoam, and then began a 3D model. Evan started working with our new linear slides (see the picture below); we expect to use linear slides a lot this year, with reaching into the craters and hooking onto the rover. We pre-drilled some holes into these new slides using an optical punch and a drill. Janavi worked with a different linear slide kit, this kit is lighter than our new kit, which is helpful, but it is very delicate and requires precision to put together.
      Evan looking through an optical punch
      Evan with a linear slide
    • Field setup
    • Many of our new recruits returned today and have continued to be active. During the week, we received the field parts, so we had them help us put it together so that they can be familiar with the field design and with certain power tools. They also helped with various devices we worked on, like the linear slides, etc.
      Field assembly progress from our new recruits.
    • Chassis testing
    • We plan to use the chassis we built this summer for preliminary autonomous testing. Janavi and Kenna got Garchomp up and running today and added a better and more secure phone holder so we can run autonomous.
    • Vision and autonomous
    • We began exploring in Open CV so that we can have a visual tool to find the minerals; the algorithms we are exploring can be used for both autonomous and tele-op. We had a discussion on our goals for vision this year, which can be found at (E-20, Vision Discussion). We also began mapping our autonomous paths to act as guides to our coders.
      Open CV progress

    Today's Member Work Log

    Team MembersTaskStart TimeDuration
    KarinaRobot build and team marker design2:004 hrs
    AbhiOpen CV2:004 hrs
    EvanPrototyping2:004 hrs
    CharlotteBlog and brainstorming2:004 hrs
    EthanWorking on blog2:004 hrs
    KennaRobot build2:004 hrs
    JustinField assembly2:004 hrs
    JanaviPrototyping2:004 hrs

    Vision Discussion

    Vision Discussion By Arjun and Abhi

    Task: Consider potential vision approaches for sampling

    Part of this year’s game requires us to be able to detect the location of minerals on the field. The main use for this is in sampling. During autonomous, we need to move only the gold mineral, without touching the silver minerals in order to earn points for sampling. There are a few ways we could be able to detect the location of the gold mineral.

    First, we could possibly use OpenCV to run transformations on the image that the camera sees. We would have to design an OpenCV pipeline which identifies yellow blobs, filters out those that aren’t minerals, and finds the centers of the blobs which are minerals. This is most likely the approach that many teams will use. The benefit of this approach is that it will be easy enough to write. However, it may not work in different lighting conditions that were not tested during the designing of the OpenCV pipeline.

    Another approach is to use Convolutional Neural Networks (CNNs) to identify the location of the gold mineral. Convolutional Neural Networks are a class of machine learning algorithms that “learn” to find patterns in images by looking at large amounts of samples. In order to develop a CNN to identify minerals, we must take lots of photos of the sampling arrangement in different arrangements (and lighting conditions), and then manually label them. Then, the algorithm will “learn” how to differentiate gold minerals from other objects on the field. A CNN should be able to work in many different lighting conditions, however, it is also more difficult to write.

    Next Steps

    As of now, Iron Reign is going to attempt both methods of classification and compare their performance.

    CNN Training

    CNN Training By Arjun and Abhi

    Task: Capture training data for a Convolutional Neural Network

    In order to train a Convolutional Neural Network, we need a whole bunch of training images. So we got out into the field, and took 125 photos of the sampling setup in different positions and angles. Our next step is to label the gold minerals in all of these photos, so that we can train a Convolutional Neural Network to label the gold minerals by learning from the patterns of the training data.

    Next Steps

    Next, we will go through and designate gold minerals. In addition, we must create a program to process these.

    Chassis Brainstorming

    Chassis Brainstorming By Ethan and Evan

    Task: Brainstorm chassis designs

    At the moment, we've used the same chassis base for three years, a basic mechanum base with large wheels. However, we don't really want to do the same this year. At the time, it was impressive, and not many teams used mechanum wheels, but now, its a little overdone.

    Thus, we have BigWheel. We used this as a practice design, but we ended up really liking it. It starts off with two large rubber wheels, approx. eight inches in diameter, mounted at the back and sides of the robot. Then, we have two geared-up motors attached to the motors for extra torque and power. In the front, we have a single omniwheel that allows our robot to turn well.

    Proposed Additions

    First, we need to add an intake system. For this, we're considering a tension-loaded carwash that can spring out over the crater wall. It'll pull elements in and sort them through our intake using our separator, which we will detail in a later post. Then, the robot will drive over to the lander and lift itself up. Since the main segment of the robot is based off of two wheels, we're attaching a telescoping slide that pushes off of the ground at the opposite end and pivots the front of the robot upwards. Then, the intake will launch upwards, depositing the elements in the launcher.

    Next Steps

    We need to create a proof-of-concept for this idea, and we'd like to create a 3D model before we go further.

    Meeting Log

    Meeting Log September 22, 2018 By Charlotte, Janavi, Evan, Abhi, Justin, Ethan, Arjun, Karina, and Kenna

    Meeting Log September 22, 2018

    Home Depot Trip!

    Today's Meet Objectives

    As we are starting to make more serious strides in our robot and strategy, we wish to start passing down knowledge to our new recruits. Today, we are going to continue prototyping with grabbers and various linear slide kits and we need to discuss strategy and organization for this season.

    Today's Meet Log

    • Robot strategy discussion
    • Today we have discussed more about what we want our strategy to look like. An option we are heavily considering is having a non-moving robot, in the sense that our robot is stationary and all game actions are performed using extensions from the robot, using linear slides, etc. We have discussed what game rules we need to consider, like what "parking" consists of during autonomous. For further information, see (E-34, Another Design Bites the Dust).
    • Chassis brainstorming
    • We discussed the chassis design we plan to use this season, and we decided experiment with the BigWheel chassis we build this summer. For more details on this discussion, see (E-23, Chassis Brainstorming).
    • Sorter prototyping
    • We have continued prototyping various grabbing mechanisms with sorting ability, one passive and one active sorter. The passive version we modeled in Creo and printed before practice, and the active was modeled using Legos! Our new recruits have been helping us prototype also, as we have been making a version 2 for the active model.
      Passive model
      Active model
    • New chop saw!
    • Some of the materials we are working with require power tools that we don't have or were damaged by rain. One of the linear slide kits we are working with is stainless steel, which requires a chop saw which we didn't have. We made a trip to Home Depot and bought one.
      Chopsaw in action

    • Finishing field assembly
    • Our new recruits finished up the field today. They ran into some problems along the way, including difficulty with putting on the top part of the lander, improper placement of the wing nuts, alignment of the lander in the foam tiles, and more but were able to overcome these difficulties and yielding a field for practice.
      Our freshman recruits!
    • Linear slide assembly
    • Evan and Janavi finished assembling the linear slides they were working on last week. As we build a chassis (or a wheel-less chassis) we are going to try both types to see how the weight, strength, friction, string tension, and other factors affect our gameplay. A side-by-side comparison of our linear slides cam be found at (E-61, Selecting Linear Slides)

      Battle of the Slides
    • Team marker
    • Karina narrowed down the ideas for a marker and she, with Kenna, has began building it. More about our marker can be found at (E-33, Team Marker Fun).
    • Open CV and our CNN
    • While we are waiting to begin code, we are testing many algorithms in Open CV, so we can accurately and consistently detect field minerals. These algorithms consider shape and color to map points to predict the location of the minerals. While developing Open CV, we have begun the development of a Convolutional Neural Network. Detail of our CNN training can be found at (E-22, CNN Training).
    • Location sensor
    • Today, Justin worked on making the location sensor (our fail-safe in case our encoders fail) smaller and more lightweight to help us meet with this year's size requirements (something we have had trouble with in the past).
    • Chassis testing
    • We tested the different chassis we build this summer on the field to see how they interact with the terrain (aka the crater). We found that Big Wheel was too long and didn't go over the crater at all unless it was backwards and got a running start. Garchomp (with Mechanums) went over the craters fine.

    Today's Member Work Log

    Team MembersTaskStart TimeDuration
    KarinaRobot build and team marker design2:004 hrs
    AbhiOpen CV and build2:004 hrs
    EvanBuild2:004 hrs
    CharlotteBlog and brainstorming2:004 hrs
    EthanWorking on blog2:004 hrs
    KennaRobot build2:004 hrs
    JustinBuild and field assembly2:004 hrs
    JanaviBuild2:004 hrs
    ArjunCode and blog2:004 hrs

    Autonomous Path Planning

    Autonomous Path Planning By Abhi

    Task: Map Autonomous paths

    With the high point potential available in this year's autonomous it is essential to create autonomous paths right now. This year's auto is more complicated due to potential collisions with alliance partners in addition to an unknown period of time spend delatching from the lander. To address both these concerns, I developed 4 autonomous paths we will investigate with to use during competition.

    When making auto paths, there are some things to consider. One, the field is the exact same for both red and blue alliance, meaning we don't need to rewrite the code to act on the other side of the field. Second, we have to account for our alliance partner's autonomous if they have one and need to adapt to their path so we don't crash them. Third, we have to avoid the other alliance's bots to avoid penalties. There are no explicit boundaries this year for auto but if we somehow interrupt the opponent's auto we get heavily penalized. Now, with these in mind, lets look at these paths.

    This path plan is the simplest of all autonomi. I assume that our alliance partner has an autonomous and our robot only takes care of half the functions. It starts with a simple detaching from the lander, sampling the proper mineral, deploying the team marker, and parking in the crater. The reason I chose the opposite crater instead of the one on our nearside was that it was shorter distance and less chance to mess with our alliance partner. The issue with this plan is that it may interfere with the opponent's autonomous but if we drive strategically hugging the wall, we shouldn't have issues.

    This path is also a "simple" path but is obviously complicated. The issue is that the team marker depot is not on the same side as the lander, forcing us to drive all the way down and back to park in the crater. I could also change this one to go to the opposite crater but that may interfere with our alliance partner's code.

    This is one of the autonomi that assumes our alliance partners don't have an autonomous and is built for multi-functionality. The time restriction makes this autonomous unlikely but it is still nice to plan out a path for it.

    This is also one of the autonomi that assumes our alliance partners don't have an autonomous. This is the simpler one of the methods but still has the same restrictions

    Next Steps

    Although its great to think these paths will actually work out in the end, we might need to change them a lot. With potential collisions with alliance partners and opponents, we might need a drop down menu of sorts on the driver station that can let us put together a lot of different pieces so we can pick and choose the auto plan. Maybe we could even draw out the path in init. All this is only at the speculation stage right now.

    Hanging Hook Prototype

    Hanging Hook Prototype By Abhi, Ethan, Justin, and Janavi

    Task: Design a hook for pulling the robot on the lander

    To get a head-start on latching and delatching from the lander during autonomous, we got a head start and made some hook prototypes. If your robot can just do these two things, you can score 80 points. When making this hook, it needs to be modular enough to not require much accuracy but also needs to be strong enough to hold 42 pounds. This hook works just that way.

    We designed this hook to have a slanted top to glide the robot into position if we aren't in the right place, making it very modular. In addition, we 3D printed this hook with ~80% infill in nylon after designing in PTC Creo. First, we tested it by hanging ~20 lbs of material off of it for one minute. This worked, but a little too well. While the nylon piece remained undamaged, the metal bracket it was supported by bent at a ninety degree angle. So, we had to pursue further testing.

    For our next test, we plan to hang a mass outside for a week. Dallas weather has been extreme lately, with a lot of rain, humidity, and heat. This will be the ultimate stress test; if one of our pieces can survive the outdoors, it can survive just about anything.

    Next Steps

    We're probably going to have to reprint this to be a bit more fitting for our robot, but its a good start and it works great so far.

    Meeting Log

    Meeting Log September 28, 2018 By Charlotte, Karina, Kenna, Janavi, Evan, Abhi, Justin, Ethan, and Arjun

    Meeting Log September 28, 2018

    Coding lessons with new recruits

    Today's Meet Objectives

    Since our overflow of new recruits, we have opened up two other teams 15373 and 15375, which Iron Reign will mentor and lead along with our mentorship of 3732 Imperial Robotics, who has also received new recruits. Today we plan to continue integrating them into FTC; we will begin teaching them the different expectations of an FTC team, including hard and soft skills such as coding and presenting to a panel of judges. In Iron Reign, we are going to continue prototyping various mechanisms we have designed. Also, we are going to get started with coding and autonomous.

    Today's Meet Log

    • Mentoring
    • This week, we had even more recruits join us today, so we decided to run through our Worlds presentation from last year to teach them about the judging process and our engineering process. We set their expectations for what competition day looks like, and what they need to focus on and maintain throughout the season, such as the engineering journal and outreach. We had a long discussion about subteams and we are going to let the recruits explore these subteams and decide for themselves what parts of FTC they wish to pursue.
      Presentation to recruits.
    • Linear slides
    • Janavi continued working with linear slides, which we installed on a bare chassis as well as the hook Abhi designed and printed. Near the end of practice we tested the slide and we found that it worked pretty well but we need additional tests before we can determine whether it will ba a viable option for our robot. To see more information on our linear slides, see (E-,).
    • Secret project
    • Evan worked on a secret project, details will be written about in future blog posts. See (E-34, Another Design Bites the Dust).
    • Team marker
    • Karina continued to work on our team marker. Last time we decided on the design we want to use, and she had put the idea into reality today.
      Ducky incarcerated
    • Modeling
    • Justin 3D modeled and printed wheel mounts for churros and hex shafts.
      Justin modeling
    • Replay autonomous and code mentoring
    • Over the summer, we worked on a new replay autonomous system where rather than coding an autonomous, testing it, then fixing it, we drive the robot in our intended path and that path is automatically recorded in the code. This year, we don't think that system will work, with the heavy emphasis on computer vision and the unreliable positioning of the robot after it drops off the hook on the rover. Also, today we worked with the recruits that demonstrated interest in coding. Abhi gave them a lesson and let them create their very first autonomous program by themselves (but with his guidance of course).

      Today's Member Work Log

      Team MembersTaskStart TimeDuration
      KarinaTeam marker build2:004 hrs
      AbhiCoding and teaching2:004 hrs
      EvanRobot build2:004 hrs
      CharlotteBlog and organization2:004 hrs
      EthanWorking on blog2:004 hrs
      KennaRobot build2:004 hrs
      Justin3D Modeling2:004 hrs
      JanaviRobot build2:004 hrs

    BigWheel Chassis

    BigWheel Chassis By Evan

    Task: Work on a possible chassis

    We've been toying around with the idea of using BigWheel, our Summer Chassis Research Project bot, in this year's competition with a few modifications. The idea for this robot is that it has a collection system that extends into the crater, and folds up on top of the robot. It reaches in with the collection arm, and grabs the blocks/glyphs, drives backwards and flips vertically using the drive wheels as a point of rotation. Here’s a basic sketch of what that looks like.

    The way this will be achieved is with a spring loaded lever connected to the omni wheel that makes up the holy trinity of wheels. So far I have pieced together the arm that reaches into the pit, which is powered by two NeverRest 60s and geared in a two to one ratio to significantly increase the torque. Between the two arm I plan for a horizontal beater bar to intake blocks and a slide attached to a servo to separate blocks and balls based on their size. The idea is to have a way of sorting based off of the physical shape rather than by digital sensing means. The more that can be done purely off the shape of the elements, the better.

    Next Steps

    Next week, the team will have to make some serious progress since there will be more hands to build. My hope is that the lever will come about soon, even if in its most infant stage, and that some semblance of a functioning robot can be game tested in the next few weeks, just in time for a scrimmage and potentially an early qualifier.

    CNN Training Program

    CNN Training Program By Arjun and Abhi

    Task: Designing a program to label training data for our Convolutional Neural Network

    In order to use the captured training data, we need to label it by identifying the location of the gold mineral in it. We also need to normalize it by resizing the training images to a constant size, (320x240 pixels). While we could do this by hand, it would be a pain to do so. We would have to resize each individual picture, and then identify the coordinates of the center of the gold mineral, then create a file to store the resized image and coordinates.

    Instead of doing this, we decided to write a program to do this for us. That way, we could just click on the gold mineral on the screen, and the program would do the resizing and coordinate-finding for us. Thus, the process of labeling the images will be much easier.

    Throughout the weekend, I worked on this program. The end result is shown above.

    Next Steps

    Now that the program has been developed, we need to actually use it to label the training images we have. Then, we can train the Convolutional Neural Network.

    Intake Sorter

    Intake Sorter By Abhi

    Task: Design a sorter for the balls and blocks

    To increase the efficiency of our robot, we looked into ways to passively sort minerals during intake and deposit. It is important to sort because it requires less precision under driver control allowing a faster and more efficient robot. Though bulky, we designed an initial design to sort the minerals.

    When this piece is mounted and both blocks and balls are run over it, the balls run down the top and don't fall in the collector, but the blocks fall in the holes. We modeled this design in PTC Creo, then printed it in ABS.

    Next Steps

    This design works but is large so we're going to have to find a smaller and simpler way to sort game pieces. In the future, we're going to minimize this and probably move to a smaller sorting mechanism.

    Designing Wheel Mounts

    Designing Wheel Mounts By Justin

    Task: Create wheel mounts for our Mini-Mecanum chassis

    Today, we modeled two possible designs for mini-mecanum wheel mounts. The purpose of the mounts is to hold a churro or hex shaft in place to mount mecanum wheels to. The first design was a 6cm by 6cm square with rounded edges that was 5mm thick. A hexagon was removed from the center to hold the churro that supports the mecanum wheel. This design, when printed on low infill, allowed the churro to rotate when enough force was applied. We modeled this design off of the wheel mounts on Kraken and Garchomp; the only differences are the size and material. Because we will be 3D printing these mounts, material efficiency is very important. This mount design used a lot of material to make a prototype, meaning a finished stable mount would need even more material to prevent the churro or hex shaft from slipping.

    Taking these problems into account, we designed a different way to mount the wheels. The new version can mount underneath a REV Rail and hold the shaft or churro perpendicular to the rail. This design uses much less infill than the previous one because of how small the mount is, and because the REV Rail also acts as support to prevent the churro or shaft from spinning. The mount also allows the mini-mecanum wheels to be mounted as close to the frame as possible, which can help make the robot more compact. This design will allow us to easily mount mini-mecanums to our frame, while using minimal filament and taking up very little space.


    Next Steps

    We need to build the full mini-mecanum robot to judge whether these designs will fully work.

    Designing the Corn Cob Aligner

    Designing the Corn Cob Aligner By Ethan and Abhi

    Task: Design an aligner for the beater bar intake

    The ice cube tray is 9 holes wide and each hole is 16.50mm wide and long. Using these measurements, we created an aligner that would cause the ice cube tray to roll into a cylinder.

    We're designing an intake that will allow the robot to intake particles, and this is a major portion. This will allow us to increase the amount of friction put on the particles, allowing for a more secure grip.

    However, this system has issues. First, we wanted the edges to still be mildly compliant, and this wheel filled out the edge rows to full depth, making them a little too tough. Plus, they made the silicone height too variable, so that we couldn't solely pick up the balls. So, we designed a second aligner with shorter spokes so that the edges would be fully compliant while still being held securely.

    Next Steps

    We need to finish up the corn-cob beater bar, but after that we'll be able to start testing.

    Corn-Cob Intake

    Corn-Cob Intake By Ethan and Abhi

    Task: Design an intake system unique for balls

    Right now, we're working on a static-deposit system. The first part of this system is having an intake mechanism that passively differentiates the balls and cubes, reducing complexity of other parts of the design. Thus, we created the corn-cob intake.

    First, we bought ice-cube trays. We wanted a compliant material that would grip the particles and be able to send them into a larger delivery mechanism.

    Then, we designed a wheel which' spokes would fit into the holes on an ice cube tray, allowing the tray to stay static while still being compliant in a cylindrical shape. Then, we can put axle hubs through the center of the wheel, allowing us to mount the wheels on a hexagonal shaft. Then, we can mount a sprocket on that, allowing the bar to be rotated for intake. This bar is mounted at the height of the balls, not blocks, so we can passively sort the minerals in-action.

    Next Steps

    We need to mount this on our robot and design a way to deliver the field elements. We're also going to go into more detail on the ice cube mounts in a later blog post.

    Team Marker Fun

    Team Marker Fun By Karina

    Task: Create the Team Marker

    Last week, we decided to take up the task of creating the team marker, a simple project that would surely be overlooked, but can score a significant amount of points. We wanted the marker to be meaningful to the Iron Reign, but also follow the team marker rules. To start, we made a list of ideas:

    Last season, Ducky (as seen in idea #4) brought Iron Reign good luck whenever the drivers squeezed them, and so we knew that we wanted to incorporate Ducky into whatever the final product would be. Some team members suggested fusing together multiple rubber duckies to fit the dimensions in the rule book. I had a better idea. I thought, "Why not put Ducky in a box?" However, trapping Ducky in a box would prevent us from ever squishing Ducky again (as long as they are trapped in the box). But then an even better idea came up: "Why not put Ducky in a cage?" And so we got to work making a cage for Ducky, one that we could release them from or reach in to whenever we need a squish for good luck.

    We cut two pieces of 3.5 inch x 3.5 inch polycarb to serve as the ceiling and floor of the cage. Then we used 8 standoffs, in pairs of two at each corner of the cage, to serve as the bars. To not waste anymore standoffs, we used zipties as the cage bars. Additionally, the flexibility of the zipties allow us to squeeze Ducky out of the cage from in between the bars. In the end, Ducky looked like the most happy prisoner we've ever seen:

    Next Steps

    With the team marker built, we need to test how well it does its job (staying in one piece for the duration of a match hopefully). It's survived many nights now in the our coach's house, which is no small feat, with all the children running about and constantly misplacing things. Once we have an intake system working for the minerals, we will need to test how compatible it is with Ducky in a Cage. Lastly, we need to decorate Ducky's cage, including our team's number (6832).

    Another Design Bites the Dust

    Another Design Bites the Dust By Ethan

    Task: Discuss a new rule change

    At one point, we were thinking about creating a "mining facility" robot that stays static within the crater and delivers the blocks into the mining depot. In our eyes, it was legal as it would hold as many blocks as possible inside the crater but only deliver two at a time outside. It would be super-efficient as we would be able to stay within the crater, and not need to move.

    However, fate has struck. Earlier this week, we received this message:

    The rule limiting control/possession limits of minerals has been updated to indicate that robots may _temporarily_ hold more than 2 minerals in the crater, but must shed any excess over prior to performing any other gameplay activities (which would include scoring).
    says that "Robots In a Crater are not eligible to Score Minerals". Per the definitions of "In" and "Crater", if _any_ portion of a Robot is in the vertical area above the crater (extending from the field walls to the outside edge of the Crater Rim), then scoring a Mineral results in a Major Penalty.
    says that Robots may not obstruct another Robot's path of travel in the area between the Lander and a Crater for more than 5 seconds.

    This means that we couldn't do a static mining facility as we cannot score within the crater. Since we'd have a portion of the robot always in the crater, the existence of our robot would be a major penalty.

    Next Steps

    So, we need to rethink our robot. We still want to create a semi-static robot, but we need to redesign the intake portion.

    Labelling Minerals - CNN

    Labelling Minerals - CNN By Arjun and Abhi

    Task: Label training images to train a Neural Network

    Now that we have software to make labeling the training data easier, we have to actually use it to label the training images. Abhi and I split up our training data into two halves, and we each labeled one half. Then, when we had completed the labeling, we recombined the images. The images we labeled are publicly available at https://github.com/arjvik/RoverRuckusTrainingData.

    Next Steps

    We need to actually write a Convolutional Neural Network using the training data we collected.

    Meeting Log

    Meeting Log October 06, 2018 By Charlotte, Kenna, Janavi, Ethan, and Arjun

    Meeting Log October 06, 2018

    Code Testing with Arjun

    Today's Meet Objectives

    We set up some tables with FTC Starter Kits for our new recruits so we can give them an introduction to building with REV parts. We want to continue research & design and build for Iron Reign. There is a scrimmage coming up in a few weeks, so we want to have a working chassis by then.

    Today's Meet Log

    • Chassis build
    • Kenna and Janavi worked on a chassis. We hope to mount the linear slides we completed last time onto this chassis and hopefully use it for our first scrimmage. We had a frame for the chassis done last time, and this time we added motors and one of four wheels. Hopefully, the chassis will be complete by next week and then we can run some test to determine whether or not it will be a viable chassis for competition use. If we deem that it is worthy, there are a few problems we need to fix before competition day. Notably, the chassis doesn't fit within the sizing cube, as it measures 17 in x 18 and 1/16th in. Our chassis decision process can be found at (E-16, Choosing Drive Train).

      Kenna with the chassis frame (pre-motored)

      Kenna and Janavi installing the motors
    • Engineering journal discussion
    • We discussed what we want to improve in our engineering notebook this year. In previous years, one of our greatest weaknesses has been the lack of mathematical analysis in our blog posts, so this year we are going to focus on doing more parts testing and incorporate statistics and physics from those tests into our blog posts.
    • Intake prototyping and design
    • Ethan has been working on prototyping with grabbers. Abhi designed and printed parts to mount our "corn on the cob" material, and Ethan put it together and made a small frame to put it on so we can test it. To see more about the intake aligner, see (E-31, Designing the Corn Cob Aligner). To see more about "corn on the cob," see (E-32, Corn-Cob Intake).

      Ethan working on the blog

      Ethan with the "corn on the cob"
    • Gantt Chart
    • Today, I made some real progress on our team "Gantt" chart. We hope to utilize such a chart in order to improve team organization and structure. Hopefully, this will prevent certain subteams from falling behind and we will not be rushed right before competitions as that has happened a lot historically.
    • Code testing and CNN training
    • Once he updated the FTC app, Arjun he tested our code with the new update on Kraken, our robot from last year. He also took 72 pictures of the minerals for training of a convolutional neural network. He began compiling those images and will work on the neural network in the coming weeks. See more about our CNN training process in (E-21, CNN Training)

    Today's Member Work Log

    Team MembersTaskStart TimeDuration
    CharlotteBlog and organization2:004 hrs
    EthanWorking on blog2:004 hrs
    KennaRobot build2:004 hrs
    JanaviRobot build2:004 hrs
    ArjunCode updates2:004 hrs

    Upgrading to FTC SDK version 4.0

    Upgrading to FTC SDK version 4.0 By Arjun

    Task: Upgrade our code to the latest version of the FTC SDK

    FTC recently released version 4.0 of their SDK, with initial support for external cameras, better PIDF motor control, improved wireless connectivity, new sensors, and other general improvements. Our code was based on last year's SDK version 3.7, so we needed to merge the new SDK with our repository.

    The merge was slightly difficult, as there were some issues with the Gradle build system. However, after a little fiddling with the configuration, as well as fixing some errors in the internal code we changed, we were able to successfully merge the new SDK.

    After the merge, we tested that our code still worked on Kraken, last year's competition robot. It ran with no problems.

    Mining Base 2.0

    Mining Base 2.0 By Ethan

    Task: Rethink our static robot idea

    So, our dream this year is to create a static robot. Last week, we found out about a rule change that would prevent our mining robot from staying within the crater. Naturally, we found a way around this, leading us to the Mining Base 2.0.

    The robot will be fixed under the lander's hooks, and have a horizontal and vertical linear slide attached to it. The horizontal linear slide would reach over the crater walls and pick up the silver balls, and shoot them up towards the lander. On the lander, our vertical linear slide would create a backboard that would allow the balls to fall into the lander. This wouldn't violate the rules as we wouldn't be in the crater. And, it would give us the benefit of having an extremely high-scoring robot.

    Next Steps

    We need to start on the designs of this robot, but to do this, we first need to create a working chassis.

    Project Management

    Project Management By Charlotte

    Task: Improve Iron Reign's team organization and time management

    Iron Reign sometimes struggles with our team organization and time management. There have been many instances where we have fallen behind in different subteams due to this lack of organization. This year, in order to tackle this downfall, we are going to put an emphasis on project management.

    We started a project in a program called Team Gantt. We learned how to use this program from watching the many tutorials in the program and by trial and error. In our project, we have made task groups that represent our subteams, such as build, code, etc. You can see this in the image above, but I did not include the whole chart to not expose any team secrets. A project manager will be in charge of keeping these subteams on track with the chart, and will update it accordingly along with periodic meetings regarding the chart and our progress. Hopefully, this will really help us in our team organization so that we don't fall behind this season.

    Next Steps

    Continue the use of our Gantt chart in order to improve our organization and give us a big-picture view of our progress for the rest of the season.

    BigWheel+

    BigWheel+ By Evan

    Task: Continue work on BigWheel

    BigWheel has gone through a few major changes. First and foremost, it now has a flipper arm, AKA Superman. Since the robot itself is the lift mechanism, we had to put a lot of work into Superman's design. Right now it is a 10 inch REV rail attached to two 125-tooth gears for redundancy, with a custom 3D printed mount housing an pair of omniwheels on the other end. On the motors, we have two 15-tooth gears, resulting in a 3:25 gear ratio. This gives us a ridiculous amount of torque that lifts the robot up smoothly. On top of the flipper, we’ve added extra supports on the arm mounts, as when we went to the Hendricks scrimmage, we found that the two sides were out of alignment, and one was bending more forward than the other, making the arm bend unevenly to one side and throwing the whole robot out of alignment.

    The next step is to strengthen the arm itself, as the two sides have a tendency to want to do their own things, mainly the side with the intake motor mounted to it. Since the supports have been put in though, Bigwheel has been functioning much better, and the arm no longer flops to one side. General wire management has also taken place, as we'd dealt with wires getting stuck in the gears.

    Next Steps

    Bigwheel was built on a bit of a shabby base, mostly being made of a piece of polycarb and some aluminum bars, and not giving much in terms of change. We’ve cut here and there, drilled a few holes, unattached and re-attached a couple of things, but in all it’s a very stiff robot, and doesn’t lend itself to fluidity of design. That’s why we plan on making a second version of this base, hopefully with thinner polycarb and more secure sides that have been welded together but can be removed more easily. The exact design is still being modeled, but we have a direction to jump off from, and I believe we can make that leap to a better robot.

    Developing a CNN

    Developing a CNN By Arjun and Abhi

    Task: Begin developing a Convolutional Neural Network using TensorFlow and Python

    Now that we have gathered and labeled our training data, we began writing our Convolutional Neural Network. Since Abhi had used Python and TensorFlow to write a neural network in the past during his visit to MIT over the summer, we decided to do the same now.

    After running our model, however, we noticed that it was not very accurate. Though we knew that was due to a bad choice of layer structure or hyperparameters, we were not able to determine the exact cause. (Hyperparameters are special parameters that need to be just right for the neural network to do well. If they are off, the neural network will not work well.) We fiddled with many of the hyperparameters and layer structure options, but were unable to fix the inaccuracy levels.

     1
     2
     3
     4
     5
     6
     7
     8
     9
    10
    11
    model = Sequential()
    model.add(Conv2D(64, activation="relu", input_shape=(n_rows, n_cols, 1), kernel_size=(3,3)))
    model.add(Conv2D(32, activation="relu", kernel_size=(3,3)))
    model.add(MaxPooling2D(pool_size=(8, 8), padding="same"))
    model.add(Conv2D(8, activation="tanh", kernel_size=(3,3)))
    model.add(MaxPooling2D(pool_size=(8, 8), padding="same"))
    model.add(Conv2D(4, activation="relu", kernel_size=(3,3)))
    model.add(Conv2D(4, activation="tanh", kernel_size=(1,1)))
    model.add(Flatten())
    model.add(Dense(2, activation="linear"))
    model.summary()
    

    Next Steps

    We have not fully given up, though. We plan to keep attempting to improve the accuracy of our neural network model.

    Meeting Log

    Meeting Log October 13, 2018 By Charlotte, Janavi, Ethan, Arjun, Abhi, Justin, and Karina

    Meeting Log October 13, 2018

    Sumo bots at SEM STEM Spark

    Today's Meet Objectives

    Today we are taking part in a massive outreach event to teach STEM to girls all over North Dallas: SEM STEM Spark. However, we do have competitions/scrimmages coming up really soon, so we wish to get some substantial building done. See more about the event at (T-22, SEM STEM Spark).

    Today's Meet Log

  • Chassis build
  • We scrapped the chassis we worked on last meeting because of it lack of mounting points and poor assembly. Janavi started with just some extrusion rails and mounted some motors and wheels for a new new chassis. Hopefully we will have a working chassis by the time of the scrimmage.
  • CNN Training
  • Arjun continued to work on a convolution neural network, which, once the network is complete, we will compare with Open CV. We have used Open CV for our computer vision algorithms for a couple of years, but we are now looking into other options to see if CNN will be a more accurate method of differentiating between field elements. A summary of our vision decisions can be found at (E-81, Vision Summary)
  • SEM STEM Spark outreach
  • Besides working on the chassis and a CNN, most of us taught and shared our passion for STEM at the event. The event was 10 hours long, so it was a long haul, but we had a really great time and the girls did too.

    Today's Member Work Log

    Team MembersTaskStart TimeDuration
    CharlotteOutreach8:0010 hrs
    EthanOutreach8:0010 hrs
    JanaviBuild8:0010 hrs
    ArjunConvolution Neural Network8:0010 hrs
    AbhiOutreach8:0010 hrs
    KarinaOutreach8:0010 hrs
    JustinOutreach8:0010 hrs

    Mini Mecanum Chassis

    Mini Mecanum Chassis By Janavi and Justin

    Task:

    Over the summer, we designed many robots for the North Texas Chassis Project, including one based off of last year's Worlds robot, Kraken. The robot chassis had 6" mechanums. But, based on what we know about this years challenge we have decided that this chassis does not utilize the 18-inch cube effectively.

    We have chosen to design a chassis that is similar in function to Kraken, but smaller in size with 4" mecanum wheels.

    Our plan is to design a low-lying 6" x 6" robot, a marked difference from the usual 18". However, this new design means that many of our 3D printed parts are unusable on this robot; for example, our former wheel mounts are much too large for the new robot and wheels, as well as their corresponding axles.

    These bearings are hexagonal, requiring a new wheel mount design.

    Justin first designed the axle plate below to solve this, but it raised the robot off the ground quite a bit, risking debris becoming stuck under the bot. As well, it was flimsy - it was mounted too far from the robot. We went back to the drawing board and brainstormed various methods we could use to attach the axle the frame in a more secure way; we found that we use a pillow block design would save space, while also having a lower-lying robot. This design worked out beautifully, leading to the design we are currently using.

    The axles and wheels aren’t the only new thing about our robot: we've switched to NeverRest 20s in lieu of our normal 40s and 60s. This is another reason that we wanted to create such a minute robot. The gear ratio combined with the size will make this robot a speed demon on the field and allows us to dart between the minerals and the depositing location quickly.

    Next Steps

    In the upcoming weeks we will continue to tinker with this chassis design by adding a linear side and our gathering mechanism, and hopefully, we will be able to demonstrate it at the scrimmage next week.

    Rewriting CNN

    Rewriting CNN By Arjun and Abhi

    Task: Begin rewriting the Convolutional Neural Network using Java and DL4J

    While we were using Python and TensorFlow to train our convolutional neural network, we decided to attempt writing this in Java, as the code for our robot is entirely in Java, and before we can use our neural network, it must be written in Java.

    We also decided to try using DL4J, a competing library to TensorFlow, to write our neural network, to determine if it was easier to write a neural network using DL4J or TensorFlow. We found that both DL4J and TensorFlow were similarly easy to use, and while each had a different style, code written using both were equally easy to read and maintain.

     1
     2
     3
     4
     5
     6
     7
     8
     9
    10
    11
    12
    13
    14
    15
    16
    17
    18
    19
    20
    21
    22
    23
    24
    25
    26
    27
    28
    29
    30
    31
    32
    33
    34
    35
    36
    37
    38
    39
    40
    41
    java
    		//Download dataset
    		DataDownloader downloader = new DataDownloader();
    		File rootDir = downloader.downloadFilesFromGit("https://github.com/arjvik/RoverRuckusTrainingData.git", "data/RoverRuckusTrainingData", "TrainingData");
    		
    		//Read in dataset
    		DataSetIterator iterator = new CustomDataSetIterator(rootDir, 1);
    		
    		//Normalization
    		DataNormalization scaler = new ImagePreProcessingScaler(0, 1);
    		scaler.fit(iterator);
    		iterator.setPreProcessor(scaler);
    		
    		//Read in test dataset
    		DataSetIterator testIterator = new CustomDataSetIterator(new File(rootDir, "Test"), 1);
    			
    		//Test Normalization
    		DataNormalization testScaler = new ImagePreProcessingScaler(0, 1);
    		testScaler.fit(testIterator);
    		testIterator.setPreProcessor(testScaler);
    		
    		//Layer Configuration
    		MultiLayerConfiguration conf = new NeuralNetConfiguration.Builder()
    				.seed(SEED)
    				.l2(0.005)
    				.weightInit(WeightInit.XAVIER)
    				.list()
    				.layer(0, new ConvolutionLayer.Builder()
    						.nIn(1)
    						.kernelSize(3, 3)
    						.stride(1, 1)
    						.activation(Activation.RELU)
    						.build())
    				.layer(1, new ConvolutionLayer.Builder()
    						.nIn(1)
    						.kernelSize(3, 3)
    						.stride(1, 1)
    						.activation(Activation.RELU)
    						.build())
    				/* ...more layer code... */
    				.build();
    

    Next Steps

    We still need to attempt to to fix the inaccuracy in the predictions made by our neural network.

    Intake Update

    Intake Update By Ethan, Abhi, Justin, and Kenna

    Task: Update the intake for the new robot size

    We created the corn-cob intake a few weeks ago. Unfortunately, it was a little too big for the Minichassis, so we had to downsize. So, we designed Intake Two. Continuing our history of using kitchen materials to create robot parts, we attached two silicone oven mitts to a beater bar equipped with Iron Reign's REVolution system. Then, we attached a REV Core Hex Motor to the design, then added a 2:1 gear ratio to increase the speed, as the motor wasn't exactly what we wanted.

    Then, we attached our new passive sorting system. Instead of being the old, bulky sorting system, the new system is just three side-by-side bars spaces 68mm apart with tilted wings to move blocks upwards. The 68mm number is important - the size of a gold block. This allows the balls to be struck and fly upwards into the intake while sliding the blocks through the system.

    Next Steps

    We need to attach this to the robot to test intake. The most likely way this'll be done is through a pivot over the walls of the crater from the top of the robot.

    Meeting Log

    Meeting Log October 20, 2018 By Charlotte, Kenna, Janavi, Ethan, Arjun, Justin, and Abhi

    Meeting Log October 20, 2018

    Juggling the minerals

    Today's Meet Objectives

    Our first scrimmage is next weekend, so we need to complete our chassis and some sort of intake system. Every member needs to take on their own portion of the robot so we can divide and conquer to end today's meeting with a working robot.

    Today's Meet Log

  • Mini-Mech chassis build
  • Finally, we have a chassis. We used small mechanum wheels and a small rectangular frame which is very unusual for Iron Reign with our history of 18 in x 18 in robots. The chassis that Janavi build last weekend during the outreach event was a square, but we needed to make it rectangular to make room for motors. See more on mini-mech at (E-42, Mini Mechanum Chassis).
  • Linear slide build
  • Janavi and Justin worked on the linear slides that Janavi has been working on for a few weeks. Before, we had tested and mounted the slide to an existing chassis, but there were some improvements to be made. They changed the length of the linear slide from using 18 in rails to 12 in rails and added stops so that the slide don't slide out of each other. They also strung the slides so that they can extend and retract depending on the direction of rotation of the wheels.

    Janavi, Justin, and some slides
  • Code mentorship
  • Arjun worked with a few members from Iron Star and Iron Core so that they could start programs for the robots they have been working on. A few weeks ago, Abhi gave them an introduction to coding, but Arjun helped them from the very beginning of making a new project and writing their first lines of code. Iron Reign has been utilizing GitHub for many years and we have found it very helpful, so we helped the other teams set up their own GitHub repositories and taught them how to use it.

    Arjun and the phone mount

    Teaching freshmen GitHub
  • Intake system build
  • Ethan and Abhi worked on our intake system. We are using silicone mats for kitchen counters to launch field elements into our intake system. The minerals then are filtered through 3 bars, each space by 68 mm so that balls roll over and cubes fall in. They completed the intake mechanism, but their greatest challenge is fine tuning the sorting bars and finding a way to mount it onto the chassis. Eventually, we wish to make the system pivotable, but for now we mounted it to the chassis so that it is stationary. Details about this intake system can be found at (E-44, Intake Update).

    Intake mechanism with red silicon mats

    Today's Member Work Log

    Team MembersTaskStart TimeDuration
    CharlotteBlog and organization2:004 hrs
    EthanWorking on blog and intake build2:004 hrs
    KennaRobot build2:004 hrs
    JanaviLinear slide and chassis build2:004 hrs
    ArjunBuild and mentoring2:004 hrs
    KarinaRobot Build2:004 hrs
    AbhiIntake Build2:004 hrs

    Off-Schedule Meeting Log

    Off-Schedule Meeting Log October-2 23-2, 2018-2 to October 23, 2018 By Ethan, Karina, Charlotte, Kenna, Arjun, and Evan

    Meeting Log October 21 to October 23, 2018

    Iron Reign will be attending a scrimmage on Saturday, but to attend a scrimmage, you usually have to have a working robot. As of Saturday, we did not. So, a few of our members elected to come in on Saturday to do some last minute robot additions.

    Sunday Tasks

    • Attached lift
    • We've had a linear slide that we've been meaning to hook up to the robot for awhile, and we finally did it Saturday. We mounted it to the front of the robot, as it was the easiest access point, then mounted a motor and pulley on the side to extend it. It worked - and then it didn't - as it tangled itself inside the motor, necessitating a redesign.
      Then we realized a more pressing issue. Since torque is equal to force * arm length (T=FR), and the force on our robot is only the force due to gravity (F=mg), we had a torque on the lift equal to T=mgR. Then, as the lift was mounted at the very end, the torque on the arm was at its absolute maximum. And, while we're confident in our building ability, we're not that confident. So, we realized that we'd have to move the lift closer to the middle to minimize torque.
    • Finished intake
    • On Saturday, we worked on the red-silicone intake system, but there were still issues. We used too-long screws to mount the motor that cut into the sprocket, we mounted the fins a little to far out so that the silicone was running into them and losing energy, and we didn't have a way to mount it. First, we replaced the 15mm M3 screws with 8mm ones, ensuring that there would be no further collision. Then, we removed the beams the fins were mounted on and replaced them with a simple crossbar the we directly mounted the fins to. That way, we could adjust all of the fins at once instead of individually cutting each beam.
    • Second stage
    • Our robot is a little on the small side for Iron Reign. To mitigate that, we planned to add a second stage to the robot for support and to hold components like the second REV hub. So, we started on that, cutting the standoffs, and attaching one side completely so that we could use it as a proto-phone-mount.

    Monday Tasks

    • Moved lift
    • To minimize torque, we moved the lift to the center of the robot. Now, this won't eliminate the torque - one side of the robot is much heavier than the other, but it makes it much more manageable.
    • Mounted intake
    • To have a functional robot, we have to have an intake *on the robot*. We had an intake, but it certainly wasn't anywhere close to being on the robot. So, we mounted a Core Hex Motor to the inside of our robot, attached a gear to our robot then bolted a second gear to our intake. Then, we attached the gear to a churro rail mounted on the robot and moved the motor to where the gears coincided. Originally, we planned to use a 30->90 gear system for a 1:3 gear ratio for a calculated 9.6 Newton-meters of torque, but this systed wouldn't fit within the size constraints, so we had to settle for a 1:1 ratio at 3.2 N*m.
    • Mounted 2nd arm
    • On our other robot, Bigwheel, we mounted the 2nd arm for a future beater bar. Unlike most of our robots, this one is mostly off-the-shelf, with some additional Textrix parts and a REV hub.

    Tuesday Tasks

    • Finished 2nd stage
    • To be able to support our additional motors, we had to add a second REV hub. And, to do that, we had to finish the 2nd stage. This wasn't that difficult, all we had to do was attach a standard piece of REV extrusion to the remaining standoffs, then add a REV hub mount, then mount the actual hub.
    • Reinforced lift
    • Our lift is a little bit wobbly laterally, so we took steps to fix this. We attached a small piece of REV rail to the second stage from the lift to minimize wobbling. This still needs to be worked on, as the rail isn't mounted well, but we'll burn that bridge when we get to it.
    • Strung lift
    • Since our lift needs to extend and retract reliably, we have to use a double-pulley system. So, we strung upwards normally, but then attached another string to a higher up pulley that could pull the whole system back down.
    • Replaced lift motor
    • Our old pulley-motor was an AndyMark Neverrest 60. Now, we have nothing against these motors, but we wanted something that would be easier to connect to the REV hub. So, we replaced it with a HD Hex Motor with a 40:1 gearbox. This actually increased the torque by a negligible amount (from 4.186 N*m to 4.2 N*m), and was a more convenient change.
    • Added scoring box
    • Originally, we cut a box template out of polycarb that was the exact size of two silver particles. Unfortunately, we couldn't find a heat gun, so we had to go back to cardboard.
    • Added intake bar
    • We added the corn-cob intake from a few weeks ago onto this robot so that it can get both blocks and balls from over the crater wall.

    Now, in theory, we have a competition-ready robot.

    Before

    After

    Next Steps

    We still need to program our robot and fix any gremlins that pop up; this will happen at the Friday meet.

    DISD Scrimmage at Hedrick MS

    DISD Scrimmage at Hedrick MS By Charlotte, Janavi, Ethan, Evan, Justin, Karina, and Abhi

    Task: Compete at the Hedrick MS DISD Scrimmage

    Today, Iron Reign competed in the DISD scrimmage at Hedrick Middle School. This was the first scrimmage of the year, so experienced teams and rookie teams alike struggled to get a working robot on the field. We go to this scrimmage every year, and it helps us gage just how much needs to be done to have a qualifier-ready robot. This year, that is a lot. We actually had two robots relatively pieced together, a main chassis and a backup, but we didn't account for many different problems that rendered them inoperable. In the case of the backup robot, the linear slide fell apart easily and was threaded so that it could only extend, and not retract. In the case of the actual robot, most of our problems stemmed from the intake system. Since we built it so recently, we were never able to write any code until in the final few days of preparation. We weren't able to debug the code and it has caused many complications in our robot. Our drive train also had many issues which we have been trying to fix and fine tune.

    Due to these many issues, we did not compete for most of our matches. We spent a lot of time working on our bots and talking to other teams about their progress and plans for the season, as well as see all of the interesting ideas they have put together in fruition in a game setting. In the match we did compete in, we did very badly due to driver error and mechanical errors in the drive train.

    BigWheel Arm

    BigWheel Arm By Evan

    Task: Design an arm for BigWheel

    Bigwheel’s intake arm is one of the most important parts of the robot. Since our scrimmage, we have learned how to make this arm much more efficient, starting with some supports. The original intake arm was made of two scrap Tetrix rails. The result of this was that the two sides of the arm would be out of sync, creating a twist in the arm that caused it to move oddly. Thus, it has been stabilized with cross beam REV rails.

    The next upgrade on the arm is going to be the box to hold the minerals. Right now it’s just a cardboard prototype and we need to move to the next version. After a bit of debate, we decided to craft it out of polycarb. The reason polycarb was not our immediate solution is because it’s unfortunately quite heavy, and instead the first thing we came to think of was thin plywood and duct tape. Thin slices of plywood would be taped together to create a fabric like box that still had form. This idea still lent itself to breakage, and we next went to a design using a thin plastic sheet, the same kind of plastic that is used inside milk cartons. The only issue is that it’s super weak and doesn’t form well, so we would have to build a frame for it, much like the plywood and tape.

    Next Steps

    Right now we’re toying around with the idea of an arm that not only flips out but also extends using a gear and tooth track made from Tetrix parts of days gone by. The reason for this is to gain a little extra height that we were lacking before in the robot and a little more flexibility when we grab minerals from the crater. To do this I had to take apart the arm from our first ever FTC robot, and use the toothed track and gear plus the extra long tetrix bars to create the slides. So far the slides are surprisingly smooth and we have high hopes for the future of the arm.

    Full Circle

    Full Circle By Evan

    A reflection on my time at Iron Reign

    In 2012 I began competing in FTC. That year our team built a robot with a giant central arm on top of a six wheeled drivetrain that sported a ring bucket that the rings would slot into one or two at a time. The idea was that we would go bit by bit, slowly moving the rings onto the rack in the middle. This was a mediocre idea in theory, but an even worse one in practice. I think in that entire season, we only were able to score one ring, and it was when I was by myself on a practice field before a match. The whole season had led up until that moment. It was the year I learned how to wire things, how to solder wires, how to use a bandsaw, a table saw, a miter saw, and how to really think about the real world applications of what I was doing. When I scored that ring, I was so happy. I told the whole team because this is what we had been trying to do for three months without success. We never scored another ring that season, despite being in first or second place at our qualifier (which is really just a testament to how heavily you can be carried in FTC). Since then i’ve worked on, designed, and built numerous competition robots, making a smooth transition from FLL to FTC, and i’ve been there for basically every major moment in our team’s history, from the very first meeting at the Virani household to our trip to the World championship competition in Houston where we won the Motivate award. I felt the same walking up on that stage and accepting the motivate with my team as I did back in 2012 scoring that one ring. That feeling of success and pride in my work. That’s why I keep doing FTC.

    I say all of this because today I had to take apart the arm of the first robot I ever built, and I thought it was a little poetic how I was using the robot I helped build in the my first season of FTC as part of the robot in my last season of FTC. It was weird. I don’t know. It was one of those rare full circle moments that you only ever get a few of and half the time you don’t even recognize them when they’re happening and never really get to appreciate them. It really just made me think back on all my years of robotics.

    Meeting Log

    Meeting Log November 03, 2018 By Ethan, Charlotte, Evan, Janavi, Kenna, Karina, Justin, Arjun, Abhi, and Bhanaviya

    Meeting Log November 03, 2018

    Today's Meet Objectives

    So, we have one week before our first tournament. This isn't great. As you can see on our last blog post, we didn't do amazingly at the scrimmage. So, we have a lot of work to do.

    Today's Meet Log

    First and foremost, we have to work on our presentation. So, we did an hour-long presentation runthrough to ensure all team members had the content down.

    Also necessary for a good tournament is the journal. We've had a consistent 10-20 post backlog since the season started, and we've finally started cutting into it. At my current count, we're down to 7 posts left. So, we're making considerable progress on this front. Ethan already finished our strategic plan earlier this week, so all we have left is to write the blurbs and retag our posts, something we'll do on Monday.

    Finally, in order to compete, we have to have a robot. Now, we have a robot, but it isn't really working. So, Evan and Karina worked on mounting an intake system, as well as reinforcing the center lever. This will ensure that the robot can actually score by the tournament.

    On the code side, Abhi found the coefficients for PID so that he can start autonomous. As well, he started merging SDK 4.2 with our 15k-line base of legacy code so that we can take advantage of TensorFlow. On that note, we discovered that SDK 4.2 comes with mineral detection out of the box with TensorFlow - something that we've been working on since kickoff.

    Finally, we have some good news. Iron Reign has official adopted its first new member of the season: Bhanaviya Venkat. Stay tuned for her first blog post later this week.

    Today's Work Log

    Team MembersTaskStart TimeDuration
    EthanPresentation\Journal2:004 hrs
    CharlotteBlog Backlog2:004 hrs
    KennaBlog Backlog2:004 hrs
    JanaviBigWheel Arm2:004 hrs
    ArjunBlog Backlog2:004 hrs
    KarinaBigWheel2:004 hrs
    AbhiAutonomous2:004 hrs
    EvanBlog Backlog2:004 hrs
    Justin3D Modelling2:004 hrs
    BhanviyaOnboarding2:004 hrs

    Pose BigWheel

    Pose BigWheel By Abhi

    Task: New Pose for Big Wheel robot

    Historically, Iron Reign has used a class called "Pose" to control all the hardware mapping of our robot instead of putting it directly into our opmodes. This has created cleaner code and smoother integration with our crazy functions. However, we used the same Pose for the past two years since both had an almost identical drive base. Since there wasn't a viable differential drive Pose in the past, I made a new one using inspiration from the mecanum one. Pose will be used from this point onwards in our code to setup.

    We start with initializing everything including PID constants and all our motors/sensors. I will skip all this for this post since this is repetitive in all team code.

    In the init, I made the hardware mapping for the motors we have on BigWheel right now. Other functions will come in later.

    Here is where a lot of the work happens. This is what allows our robot to move accurately using IMU and encoder values.

    There are a lot of other methods beyond these but there is just a lot of technical math behind them with trigonometry. I won't bore you with the details but our code is open source so you can find the necessary help if you just look at our github!

    Torque Calculations

    Torque Calculations By Karina

    Task: Calculate the torque needed to lift chassis

    After seeing how well the robots that could latch onto the lander performed at the scrimmage, Iron Reign knew that we had to be able to score these points. We originally tried lifting with a linear slide system on MiniMech, but it was not strong or sturdy enough for the small chassis, and would definitely not be a functional system on BigWheel in time for competition. And so we thought why not use this opportunity to *flex* on the other teams with an alternative design? An idea was born.

    We decided we would latch onto the lander using the same arm used for intake, and then pivot the main body of BigWheel up off of the ground about an "elbow joint", much like how humans do bicep curls. To do so, our motors would need to have enough torque to be able to lift the loaded chassis off the ground once the arm hooked onto the latch. First, we measured the mass of BigWheel. Then we found where the center of mass was located. The distance from the pivot point to the center of mass became our lever arm, also known as the radius.

    Calculating torque required knowing the forces acting on BigWheel at its center of mass. In this case, there was only the force due to gravity (F = mg). Before we could plug BigWheel's mass into the equation, we converted to units of kilograms (kg), and then used the value to find the newtons of force that would oppose the upward motion:

    Finally, we plugged the force and radius into the torque equation:

    Next Steps

    The next step is to test which gear train will output this torque value based on the motors used and the gear ratio.

    Linear Slide Lift

    Linear Slide Lift By Janavi

    Task: Design a lift for MiniChassis

    For extension both into the crater and lifting our robot up to the crater we have decided test a linear slide system. We plan to utilize linear slide system for both vertical and horizontal extension on MiniMech.

    Horizontal Extension Goals

    • Long Enough to reach Crater from distance
    • We need to determine how many stages we need

    Vertical Extension Goals

    • Long Enough to reach lander
    • Strong enough to support robot weight

    When designing a lift we need to determine the optimal gear ratio to allow our lift system to lift the robot but still do it relatively fast. Realistically looking at the aluminum parts we are using we plan for the robot to be around 35 lbs. We also know that the lander is 22 inches above the ground and we plan for the linear slide to extend to 14 inches off the ground This would mean that the point of rotation for our hook mechanism would be 22 inches - 14 inches = 7 inches below the latch on the lander.

    We plan to use REV 40:1 motors that have 594.7 oz*in. Now using these calculations we can determine our needed gear ratio.

    This gear ratio of 6.6 means that for our robot we need a motor to gear ratio that needs around seven rotations of the motor to provide one rotation of the hook.

    We knew the max weight of the robot would be around 20 pounds since the total weight of all the parts in the kit is less than 20 pounds. The point of rotation for the hook would be around 5.5 inches below the lander latch. This is because the bottom of the hook is around 22 inches above the ground and the point of rotation will be around 16.5 inches off the ground so that we can account for space for a gear while staying within the 18 inch box. Below is the torque calculation.

    Next Steps

    RIP CNN

    RIP CNN By Abhi

    Task: Farewell Iron Reign's CNN

    FTC released new code to support Tensorflow and automatically detect minerals with the model they trained. Unfortunately, all of our CNN work was undercut by this update. The silver lining is that we have done enough research into how CNN's work and it will allow us to understand the mind of the FTC app better. In addition, we may retrain this model if we feel it doesn't work well. But now, it is time to bid farewell to our CNN.

    Next Steps

    From this point, we will further analyze the CNN to determine its ability to detect the minerals. At the same time, we will also look into OpenCV detection.

    BigWheel Upgrades

    BigWheel Upgrades By Evan

    Task: Get BigWheel ready for the tournament

    Today, we built mounts to attach both types of intake to the rack; the rack-and-pinion corncob intake and the red-flapped intake. We also created a new way of mounting the arm to the chassis. The idea is that since it’s attached to the rack and pinion track, it reaches high enough for the robot to put the minerals in the lander. We made the arm with a 12-86 gear ratio. Our next plan is to create the mount, minimizing the size of the arm.

    The final addition is a tail for the robot to be able to stop itself from flipping backwards, something that is a very real danger of the design. It will probably be made of polycarb with aluminum or steel support on either side, just in case the polycarb is not enough to support the push of the robot. Part of this process will involve some code tuning so that the robot stops itself, but the tail is necessary as a preventative measure.

    Next Steps

    There’s still a lot of stuff we will have to do to prepare the robot physically for the competition this Saturday, but I believe it will get done.

    Conrad Qualifier

    Conrad Qualifier By Ethan, Charlotte, Karina, Janavi, Bhanaviya, Abhi, Arjun, Evan, and Justin

    Task: Compete at the N. TX Conrad Qualifier

    Right off of a mortifying experience at the Hendricks MS Scrimmage, in which we got the worst score at the tournament (and in the one match we did participate in, our robot broke) we walked in on shaky ground. In the week leading up to the tournament, Iron Reign worked hard, with 35 commits to the blog, and countless changes to our robot.

    Inspection

    Our robot fit well inside the sizing cube. However, we were warned for our rats' nest of wiring at the base of our robot, as well as the fact that our metal-frame base had sharp corners.

    Presentation

    We walked in, and started off out strong. Half of a good presentation is the energy, and we had more energy than some of our other presentations last year. Unfortunately, that energy petered out as we stuttered and tripped over ourselves. We got our information across, but not as well as we should have, and we didn't have enough time for questioning.

    Robot Game

    We didn't really have a working robot, but we tried our best. Unfortunately, our best wasn't great.

    Match 1

    We lost, 33-135. We deployed the wrong autonomous and couldn't drive - a total wash.

    Match 6

    We lost, 15-70. Our robot's linear slide seized up, bringing our robot outside of sizing limits, so we had to sit out the match as we hacksawed through our intake.

    Match 11

    We lost, 47-122. Our autonomous worked! (but our team marker didn't deploy).

    Match 13

    We lost, 65-196. Our robot didn't work, we just drove ourselves around aimlessly.

    Match 15

    We lost, 10-167. This time, none of our robots worked!

    In summary, a disappointing result.

    After-Judging and Awards Ceremony

    While we thought we hadn't done well in judging, we were quickly rebuffed. A good measure of judging success is if the judges come back to talk to you, and this was no exception. We had five separate groups of judges come up to us and ask us about *every* component of our team, from business, to volunteering, to code, to design. While we thought we hadn't done well in judging, we were quickly rebuffed. A good measure of judging success is if the judges come back to talk to you, and this was no exception. We had five separate groups of judges come up to us and ask us about *every* component of our team, from business, to volunteering, to code, to design.

    In the ceremony, every single member of SEM Robotics waited. Iron Star had been the 4th alliance captain; Iron Core had demonstrated gracious professionalism; Iron Reign had multiple in-depth discussions with judges; Imperial had an exceptional journal. We watched each team get nominated for awards, but only that, and fall short. In particular, Iron Reign was nominated for every award but Innovate. Then came Inspire. We heard two names echo off as nominations; neither of them SEM Robotics teams. Finally, a speech flew across the arena as Iron Reign stood for their Inspire Award.

    Next Steps

    Even though we won Inspire, we have a long way to go. We are going to compete at at least one more tournament, and don't want to completely embarrass ourselves.

    Code Post-Mortem after Conrad Qualifier

    Code Post-Mortem after Conrad Qualifier By Arjun and Abhi

    Task: Analyze code failure at Conrad Qualifier

    Iron Reign has been working hard on our robot, but despite that, we did not perform well owing to our autonomous performance.

    Our autonomous plan was fairly simple: perform sampling, deploy the team marker, then drive to the crater to park. We planned to use the built-in TensorFlow object detection for our sampling, and thus assumed that our autonomous would be fairly easy.

    On Thursday, I worked on writing a class to help us detect the location of the gold mineral using the built-in TensorFlow object detection. While testing this class, I noticed that it produced an error rather than outputting the location of the gold mineral. This error was not diagnosed until the morning of the competition.

    On Friday, Abhi worked on writing code for the driving part of the autonomous. He wrote three different autonomous routines, one for each position of the gold mineral. His code did not select the routine to use yet, leaving it open for us to connect to the TensorFlow class to determine which position the gold mineral was.

    On Saturday, the morning of the competition, we debugged the TensorFlow class that was written earlier and determined the cause of the error. We had misused the API for the TensorFlow object detection, and after we corrected that, our code didn't spit out an error anymore. Then, we realized that TensorFlow only worked at certain camera positions and angles. We then had to adjust the position of our robot on the field, so that we could.

    Our code failure was mostly due to the fact that we only started working on our autonomous two days before the competition. Next time, we plan to make our autonomous an integral part of our robot, and focus on it much earlier.

    Next Steps:

    We spend more time focusing on code and autonomous, to ensure that we enter our next competition with a fully working autonomous.

    Materials Testing Planning

    Materials Testing Planning By Ethan

    Task: Design a lab to test nylon properties

    So, Iron Reign is used to using off-the-shelf materials on our robot: silicone oven gloves, ice cube trays, nylon 3D-printed parts, and more. But, we've never actually done a thorough investigation on the durability and efficacy of these parts. Because of this, we've had some high-profile failures: our silicone blocks breaking on contact with beacons in RES-Q, our nylon sprockets wearing down in Relic Recovery, our gears grinding down in Rover Ruckus. So, we're going to do an investigation of various materials to find their on-robot properties.

    Nylon Testing

    A majority of the 3D-printed parts on BigWheel are nylon - we find it to be stronger than any other material save ABS, but much less prone to shattering. Still, we still deal with a substantial amount of wear, and we want to find the conditions under which damage happens.

    So, to start, we are printing a 4.5" x 1.5" block with a thickness of 4mm with an infill of 60% out of nylon. We chose these values as our average part is about 4mm thick, and our high-strength nylon pieces are about 60% infill. Then, we are going to test it under a variety on conditions meant to simulate stressful operation. As well, we're going to measure other values such as coefficient of friction using angle calculations.

    Silicone Testing

    Similarly, we use the silicone oven mitts on our intake; we find that they grip the particles pretty well. The main thing that we want to test is the amount of energy they have while rotating and then the amount of energy they lose upon collision. We plan to test this through video-analysis. In addition, we wish to test the coefficient of friction of the mitts to see if a better material can be found.

    Next Steps

    We are going to perform these labs so that we can compare the constants we receive to commonly accepted constants to test our accuracy.

    Conrad Qualifier Post Mortem - Short Term

    Conrad Qualifier Post Mortem - Short Term By Ethan, Bhanaviya, Janavi, Charlotte, Kenna, Arjun, Justin, Janavi, Karina, and Abhi

    Task: Analyze what went wrong at Conrad

    Iron Reign didn't necessarily have the best time at Conrad. As shown in last week's tournament post, the day had its ups and downs. Even though it was a successful tournament overall, there's much that we could do better.

    Problems:

    The Robot

    First, the robot didn't perform well. So, we're beginning our analysis from the mindset that everything must be changed.

    • The Intake
    • The intake itself had a multitude of problems. First and foremost, we actually didn't have a way to contain the particles from the intake. Being that Rover Ruckus' primary way of scoring is by depositing the particles into the lander, this was a pretty big oversight. To solve this, we plan to add a catcher at the bottom of the intake using this template.

      As well, our linear slide locked up in the middle of the tournament, preventing our intake from extending. Now, we have latches that keep the intake from retracting without human assistance.

    • Superman Arm
    • This impressed the judges a lot and was one of the more reliable parts of our robot. However, there were still issues. First and foremost, the arm became misaligned so that the gears began to grind during the judging presentation. This was an easy fix - we just adjusted a set screw - but we need a more rigorous solution. Right now, we're considering metal gears instead.

    The Presentation\Judging

    We didn't have much practice with our presentation. Some of the more major issues were slide order (~5 second gaps between people talking, stuttering due to unfamiliarity with content, and energy (a majority of the members present had held an all-nighter so we weren't really awake).

    We plan to revamp our presentation, adding to the story of BigWheel's development. Plus, we'll have all of our members in the next presentation, which'll be a major help. We need to do more practice, but that's a given.

    Another thing that we fell short on was the Innovate Award (the only award that we weren't mentioned for). A good portion of this is that the Innovate Award rubric emphasizes that the robot needs to work; ours really didn't. However, we need to take a retrospective look at our mechanism insofar that we need to show our difference between us and other robots.

    Programming

    Despite our all-nighter and prior large codebase, we were pretty short on workable code. So, while our driving worked, not much else did. We had an theoretical autonomous, but it remained only that.

    Next, we need to work on our Pose class (the one that determines the position of the robot on the field). From there, we need to add autonomous enhancements, allowing us to drive a little better. The most efficient use of our time could be put toward raising our robot to score and latch, as well as TensorFlow recognition of the minerals.

    Meeting Log

    Meeting Log November 17, 2018 By Charlotte, Karina, Kenna, Janavi, Evan, Justin, Ethan, Arjun, Bhanaviya, and Abhi

    Meeting Log November 17, 2018

    Evan working on the robot!

    Today's Meet Objectives

    We are going to discuss multiple facets of our team (presentation, engineering journal, organization, etc) with alumni Jayesh and Lin. What we hope to gain out of our conversation is an outside perspective. In addition to this conversation we wish to continue in our reflection of the tournament last weekend and preparation for our next tournament.

    Today's Meet Log

    • Organization
    • Karina and Janavi spent a large portion of practice organizing all of our parts and tools. They organized our drawers, carts, and tent. Organization has historically been a weak spot for Iron Reign, so this year we really want to crack down on that problem, as discussed in (T-13, Organization!).
    • Superman arm and wire organization
    • Evan, Kenna, Janavi, and Karina were all making improvements on our robot, notably working on problems we found at the tournament last week. These problems mostly dealt with wire organization and our superman arm. Analysis on why the superman arm broke can be found at (E-63, Code Issues Break the Superman Arm). More about how we fixed these issues can be found at (E-65, Arm Repairs).
    • Blogging mentoring
    • Also, Bhanaviya is learning to make blog posts. We showed her our blog post guides and answered any questions she had. Expect to hear from her soon.
    • Alumni Meeting and Feedback
    • The main focus of today's meeting was speaking to our alumni Jayesh and Lin who are both in their sophomore of college. They were both founding members of Iron Reign, they were in their senior year the first time we went to supers. More details on this meeting and our post-mortem can be found at (T-27, Conrad Qualifier Post Mortem - Short Term).
    • Presentation feedback
    • First we discussed our presentation lacked energy and enthusiasm, which is a common problem in our presentations. We have great enthusiasm for our work and progress, but we have trouble expressing it on early morning competition days. This could also be improved by lots and lots of practice, so we don't ever have to focus on our memorization, rather focusing on the expression of our passion for robotics.
    • Engineering journal feedback
    • Also, they provided insight on our engineering journal, which they said needs more cohesiveness between posts. This takes the form of adding links to older blog posts that reference future ones after we have written them.
    • Mentorship feedback
    • Finally, we discussed the new teams we have started, Iron Core and Iron Star, and asked for their advice on how to approach mentoring the new recruits. They told us that rather than waiting for them to seek us out, we need to seek them out, as many of the recruits don't have the confidence to approach us, since many of our team members are upperclassmen. We want to let them know that Iron Reign is here to help them in any way possible and to make our workspace one of collaboration and the transfer of ideas through the teams and grade levels.

    Today's Member Work Log

    Team MembersTaskStart TimeDuration
    KarinaOrganization and Build2:004 hrs
    AbhiConversation2:004 hrs
    EvanRobot build2:004 hrs
    CharlotteBlog and organization2:004 hrs
    EthanWorking on blog2:004 hrs
    KennaRobot build2:004 hrs
    Justin3D Modeling2:004 hrs
    JanaviOrganization and build2:004 hrs
    BhanaviyaLearning to Blog2:004 hrs

    Chassis Mark Two Planning

    Chassis Mark Two Planning By Ethan

    Task: Plan a new BigWheel chassis

    Our next tournament is a while away, in about two months. So, we have a little bit of time to redesign. And, our current chassis has plenty of faults.

    Our original BigWheel base.

    First and foremost, our chassis was built for a testing competition, not to be a full fledged competition robot. As such, it's a little lacking in features that would be normal on such a robot such as mounting points for other components, durability, and free space. So, we need a redesign that allows for greater modularity and functionality.

    We're starting from the ground up; our current base is a square metal frame with a polycarb bottom. While this is a good start, it has some issues: the base seems to be a little wobbly due to the polycarb, there's only one level of construction, so our motor mounts, REV hubs, and supports compete for space, and we have to add all the counting points ourselves.

    The main way to prevent the wobbliness is by replacing the polycarb with something sturdier, as well as not having everything simply bolted together. Thus, we're going to dive headfirst into the next step - welding. We plan to cut a base out of aluminum as well as four side plates to create a dish-like shape. Then, we plan to TIG weld these plates together (TIG welding uses a tungsten electrode in contact with two separate metal plates in combination with a filler metal that melts and joins the two plates together).

    Basic design for the newest version of BigWheel.

    Next Steps

    We plan to cut the aluminium next week, and TIG weld the pieces together the week after that. We're beginning to train a few of our members on TIG welding and we already have some of the safety equipment to do so.

    Conrad Qualifier Post Mortem - Long Term

    Conrad Qualifier Post Mortem - Long Term By Ethan

    What could have gone better?

    This is a document for analyzing what we can do better, not just what went wrong at the Conrad qualifier. The format of this will be in issue > solution format.

    Prep

    • Lack of tools and parts
      • Pack tools the week before - involves better organization overall
      • Bring failsafes & extra parts - prevents costly errors
    • Little presentation practice
      • Cut down powerpoint - optimally 8 minutes
      • More practice - seamless transition
      • Order - we need to tell a story
    • Journal prep
      • Same issue - we need to organize the journal to tell a story
      • Lack of images - backdate images in blog posts
      • Lack of diagrams - explanatory
      • Lack of continuity - link posts together to show how components of team have changed
      • Need to write real control award

    Programming

    • Autonomous
      • No autonomous - need to have functional autonomous
    • TeleOp
      • Robot easily breaks - need to create presets to prevent

    Build

    • Lift
      • Lift linear slide broke - need to redesign with new linear slides
    • Intake
      • Intake did not actually move - need to reattach motors

    Other

    • Presentation
      • Map slides to articles in journal
      • Review judging rubrics

    C.A.R.T. Bot Side Shields

    C.A.R.T. Bot Side Shields By Ethan

    Task: Design sideshields for the Townview Tournament

    Iron Reign takes pride in the Townview Tournament; we really enjoy making it a great experience for everyone. One small way we plan to improve the tournament is to turn our MXP into a robot repair shop for broken robots. In addition to this, we're turning CART Bot into an ambulance to carry broken bots that need repair. To do so, we're wiring a flashing light to the cart, as well as printing giant sideshields on either side. The shields are above.

    Friction Coefficient and Energy

    Friction Coefficient and Energy By Ethan

    Task: Measure the coefficient of friction of our oven mitt intake

    We want to measure various constants of materials on our robot. Earlier this season, we found that a nylon-mitt collision on our intake sapped the rotational energy of our intake. But, that was just a build error, easily fixable. But now, we plan to measure the energy lost from particle-mitt collisions, and the first part of this is to find the coefficient of friction of the silicone mitts.

    To measure the coefficient of friction, we first had to simplify an equation to determine what values to measure.

    From these calculations, we determined that the only factor to measure to determine the coefficient of friction between blocks and the mitts is the angle of incline. Therefore, we created a simple device to measure the angle at which slippage begins to occur.

    The angle was about 27 degrees, so the coefficient of friction is equal to arctan(27)=0.44. This is a pretty good coefficient of friction, meaning that the intake is very efficient in bringing the particles in, but it also means that the intake loses more energy on collision.

    Next Steps

    We need to measure further constants such as stretch and wear of nylon. To do so, we're printing a simple testing nylon block.

    Meeting Log

    Meeting Log December 01, 2018 By Charlotte, Ethan, Kenna, Evan, Abhi, Justin, and Bhanaviya

    Meeting Log December 01, 2018

    Today's Meet Objectives

    We plan to prepare for a few events coming up, the tournament we are going to host at Townview and our presentation to the Dallas Personal Robotics Group. As well, we plan to continue building our robot and improve on the superman arm in preparation for our next competition in January.

    Today's Meet Log

    • Hosting a qualifier
    • The Townview qualifier is coming up in just a few weeks, and we are starting to make preparations. Ethan is making a wrap for Cart Bot that emulates an ambulance, so we can stock the cart with tools and drive it around to help teams during the competition.

      Ethan designing
    • Robot materials testing
    • This year, we want to continue our materials testing in order to ensure our robot is efficient. Here is Ethan performing one of these tests, measuring the friction of different materials we might use for an intake system. Further information on the tests can be found at (E-59, Friction Coefficient and Energy).

      Materials friction testing

    • Model updates
    • Justin kept working on the 3D model, which is essential to complete as we are trying to improve the various systems on our robot, especially the Superman arm and other complicated mechanisms.
    • Blog training
    • A universal responsibility for Iron Reign members is writing blog posts. We taught Bhanaviya how to use GitHub and Notepad ++ so that she can write her own blog posts and post them to the blog.
    • DPRG prep
    • Abhi is preparing a demo in preparation for our meeting with the Dallas Personal Robotics Group (DPRG). We are going to show off our robot's computer vision capabilities and the strides we have made to train our own neural network. We expect to receive a lot of specific questions about this. Our presentation will be an hour long. To see how our presentation went, read (T-31, Presenting to the DPRG).

      Today's Work Log

      Team MembersTaskStart TimeDuration
      AbhiCode2:004
      EthanBlog & Testing2:004
      EvanBuild2:004
      CharlotteBlog2:004
      BhanaviyaBlog2:004
      KarinaBuild2:004
      JustinModelling2:004
      KennaSocial Media2:004

    Selecting Lift System

    Selecting Lift System By Janavi

    Objective: Determine the type of lift system will allow us to delatch and reach the lander

    In our past post Choosing Drive Train we decided that we will use the chassis BigWheel. After deciding the base we need to now think about the lift system that we want to use to allow us to both deposit into the lander and latch onto it. Evan and I have been experimenting with linear slides to use for our lift. I have been working on a REV linear slide lift system as referenced in the post "Linear Slide Lift". Evan has been working on a separate ball bearing linear slide. As well as these two options we are looking into past linear slides and ones that we have seen teams use in past challenges. We need to determine which of the linear slides works best based on the game requirements this season

    Linear slides needs according to game
    • Lift and lower robot from latch on lander
    • Extend out to Crater from distance to collect minerals
    • Extend out vertically to lander to deposit minerals

    What we want our linear slide to have
    • Light Weight
    • Bidirectional (Able to collect from crater and deposit)
    • Speed
    • Sturdy
    • Easy to fix and maintain in case of emergency
    • Small in size
    • Extend out to around 5 ft in height

    Linear Slide Options
    • Ball Bearing Lift
      • Heavy
      • Smooth
      • Reliable
      • Never used the before
    • Drawer Slides
      • Heavy
      • Low cost
      • Unwieldy
      • Familiar as we used them last year
    • REV Linear Slides
      • Light Weight
      • Not very reliable
      • Familiar

    Next Steps

    We need to select the best linear lift system for our chassis based on the requirements we set above.

    Linear Nylon Strength Test

    Linear Nylon Strength Test By Ethan

    Task: Measure linear nylon wear

    We've had some issues with our nylon sprockets, mainly through excessive wear and tear. So, we want to test what circumstances cause what deformation.

    Linear Deformation

    This one was simple. We printed this block with 60% infill (the highest infill we tend to use), measured its length (3.75") and hung one end from our deck. On the other end, we inserted a bar and attached 180 lbs of mass to it, then we measured its new length (3.8"). Thus, the constant of deformation is [weight]/[change in length] = 650 kg/cm. This demonstrates that linear transformation isn't Iron Reign's issue, as the highest possible weight put on any nylon piece on our robot is ~27 lbs/12.25kg.

    However, there is other damage. After testing, we found internal damage in the nylon from where it was hanging.

    Next Steps

    Next, we need to test the rotational damage that nylon incurs through friction. We plan to design a simple rotational sprocket and run it on a motor for a set amount of time and measure the wear to determine wear per unit time.

    Code Issues Break the Superman Arm

    Code Issues Break the Superman Arm By Abhi

    Task: Analyze the code issues that led to our robot breaking

    After constant use, our robot's Superman arm broke. At this point, it is important to analyze our failures. This error was not because of a build issue but rather a code and driver control issue.

    When testing, we always heard the gears grinding some times and we thought it wasn't an issue. There were instances like once when we accidentally made the robot stand up under a table. Other times, the robot would press the arm down into the foam and keep pushing when it couldn't really keep going, leading to grinding.

    Not only did the arm break but also the Superman mechanism. This broke mainly because we didn't set proper ranges of motion of the arm and the gears would grind when there was interference. Because of the damage, we can't use the existing gears.

    Next Steps

    We intend to gang up the gears and make the mesh stronger to fix the build side of things. In the code, I already added the software limits to motion so we don't have those problems anymore.

    Arm Repairs

    Arm Repairs By Evan and Abhi

    Task: Fix elbow and Superman

    This is a follow up to Post E-64, Code Issues Break the Superman Arm. We made some hustles and got them fixed. We reinforced Superman by ganging up multiple gears (as seen above) and repeated a similar process with the elbow arms. Hopefully this will make BigWheel more secure, especially with software limits in the code.

    Rotational Nylon Wear Test

    Rotational Nylon Wear Test By Ethan

    Task: Test the amount of wear on a moving nylon part over time

    After our last tournament, we noticed several 3D-printed sprockets that had worn down significantly. So, we wanted to measure how much wear one of our nylon sprockets takes per second.

    First, we printed out a model of one of the REV sprockets, using the STEP file here. We printed it with ~45% infill, our average for sprockets and other parts. Then, we attached a REV Core motor to an extrusion, then mounted the nylon sprocket on the other side. Then, we measured the length on one of the teeth. We ran the motor for 1:05:45, and then measured the length afterwards.

    So, the tooth length before was 5.3mm, and after, it was 5.23mm, for a difference of 0.07mm. Then, we ran the system for 1:05:45. This results in a wear rate of 1.77*10^5 mm/sec. So, given that we use our robot for about an hour, cumulatively, in a tournament, 0.0638mm, or 1.2% of the sprocket. This is enough to be noticeable under loose-chain conditions and indicates that we should keep extra sprockets at tournaments so that we can do a quick replacement if needed.

    Next Steps

    We plan to perform more materials testing in the future; in particular, we'd like to determine the wear rate of the regular REV sprockets as well, but this requires a more rigorous experiment.

    Selecting Intake System

    Selecting Intake System By Janavi

    Objective: Determine the type of intake system that will allow us to efficiently obtain and deposit minerals within the lander

    In our post "Selecting Lift System" we decided that the linear slide system that we will use is the MGN12H rails also referenced to as the Ball-Bearing slides. These slides while heavy provide the smoothest option. now that we have chosen the Lift system we need to determine the intake system that will allow us to take in two minerals and deposit them in the most efficient way possible. Throughout this season already we have been experimenting with different types of intake systems as seen in posts like "Pool noodle intake" and "Selective Intake" and "Scoring Mechanism"

    Intake System needs according to game
    • Collect only two minerals
    • Sort between silver and gold minerals

    What we want our linear slide to have
    • Light Weight
    • Speed of intake mechanism
    • Sturdy
    • Easy to fix and maintain in case of emergency
    • Small in size

    Passive Deposit vs Passive Intake

    Pros Cons
    Passive Deposit Faster intake Could be unreliable if not positioned correctly
    Passive Intake More accurate Harder to intake and therefore we score less

    Intake Mechanism Material / Shape

    Pros Cons
    Ice Cube Tray Compliant and smooth Not a far reach
    Surgical Tubing Farther reach Possibility of missing minerals due to sporadic behavior of surgical tubing
    Pot holder Brings in minerals Not far reach and too compliant
    Octopuckers ( from last year's season ) Experience with using material Too stiff and not far enough reach

    End of TIG Welding

    End of TIG Welding By Evan

    Task: Detail TIG welding plans and why they failed

    At the beginning of the season, we saw that our robot base was not as well crafted as we originally thought it to be. While we have worked to correct it over the season, it’s still not what we wish to see in a functional robot, and we came up with the idea of making the frame from light aluminum instead of the polycarb, and fix it with TIG welding.

    It seemed like a good idea at the time, but there were many other problems on the robot more important than a new base. So we pushed the TIG plan to the side, in lieu of correcting other issues like the lift and the intake. While we won’t completely throw the idea out, it will be a while before we begin to start the project. Also hindering us is the amperage output of the home, which is too low to run the TIG welder off of. Until we get additional amperage to the house, our plans will be on hold but not forgotten.

    The Return of BatteryBox

    The Return of BatteryBox By Ethan

    Task: Create a charging station for our phones and batteries

    A long time ago, in a land far, far away, Iron Reign once had a battery box. This was a fabled land, where all batteries remained charged and phones roamed the land, happy and content with their engorged batteries. But, this land was neglected, with the meadows of electricity growing dim, the plastic of the land cracking and scattering to the four corners of the Earth, and those who found their home there lost to the void.

    We have a problem keeping our phones charged at tournaments and in practice. So, we made a simple battery box to fix it. We used an old REV container and cut some spare wood to create dividers, cut a hole for a surge protector, and we were a go.

    Next Steps

    Iron Reign really needs to work on its organization in general, and this was just one way to stem the tide of entropy. We need to revitalize our tournament kits of tools next.

    BigWheel Arm Locks

    BigWheel Arm Locks By Evan

    Task: Create locks to keep BigWheel's intake arms in an extended position

    An important part of this year's challenge is scoring minerals in the lander. Additionally, our upright elbow cannot raise the scoring mechanism to the lip of the lander alone. Thus, we had to create a way to get those additional inches to score.

    First, we tried a REV linear slide design. This worked, but barely. It repeatedly got stuck, at one point even needing to be sawed apart at a tournament due to its inoperability. So, we switched to a new brand of linear slides, the MGN12H with 12mm steel rails. But, since we were no longer using REV, we needed a new design to keep the arms in the extended position.

    The new design relies on gravity. When the robot is on the lander in the hanging position, it will stay within the sizing cube. However, as it descends, the linear slides will glide upward, staying attached to the lander until the robot contacts the mat. And, as the slide finishes moving, it will move over a triangular piece of polycarb such that it is easy for the slide to move up, but near impossible to reverse its direction. This will ensure that the robot's arm stays extended.

    Next Steps

    We need to reattach the mounting point for hanging in order for this system to work.

    Scoring Mechanism

    Scoring Mechanism By Janavi and Abhi

    Task: Create a way to hold minerals

    We now have a lift and an intake system, but we're missing a way to hold onto and deposit the minerals after intake. To achieve this, we created a prototype.

    We wanted to create a box-like shape that can be attached to a moving axle to hold the minerals when lowered. When the lift is up in the air, the axle can rotate to lower the box and let the minerals fall into the depot. We tested out multiple designs but we ended up having to nix that as there was no way to get the minerals out of the box once they were in.

    Our second design was a sloped shape: just steep enough to hold onto the balls but not so steep that the balls couldn’t escape. To create this shape, we decided to have a rectangle attached to an axle able to hold the minerals when down and deposit the minerals when spun. We created multiple variations with different sizes as can be seen in the drawing below. We eventually settled on was design B, a square that was 155 cm by 155 cm.

    We decided to not use design A as it was simply too large and continuously hit against the edge of the rail. We progressed to a smaller size of 155 mm by 155 mm (design B) that worked well. We attempted another design of two separate backs as two separate channels for the minerals (Design C). However, we decided this wasn't a very good design because there was an increased chance of the ball getting stuck in between the channels, causing either a penalty or a decrease in the number of balls we can control.

    After creating the back of our holder, we realized that we needed to elevate it off the back of the rails at an angle. It was the only way to hold the balls and allow them to come down a ramp when the axle is spun. We decided that the best way to achieve this was through two wing-like triangles on the side that we could bend to ensure the minerals couldn’t escape out the side. We went through multiple designs as can be seen below

    At first, we attempted to attach two right angle triangles with 155mm acting as a leg of the triangle. We varied this design by increasing the angle of the slope so that the balls would be held at an angle that allows them to not slip. But, after creating this design out of cardboard and attaching it to the axle, we saw that the sharp angle interfered with the beater bar. To amend this, we changed the triangle attached to the end of the rectangle to have the 155 mm side be the hypotenuse of the triangle. Again, we varied the design in the steepness of the triangle. Through this, we determined that a slope of around 30 degrees was the best design.

    After finalizing our design, creating it out of cardboard, and attaching it to our robot, we cut the piece out of polycarb. We bent the side triangle using heat and drilled in the holes to attach the axle with.

    Next Steps

    Although this design works well, we want to continue to change and improve upon it. For example, the next way we can improve the design is by changing the way that the polycarb is attached to the axle through a 3-D printed attachment.

    Meeting Log

    Meeting Log December 22, 2018 By Charlotte, Ethan, Janavi, Bhanaviya, Evan, Arjun, and Abhi

    Meeting Log December 22, 2018

    Today's Meet Objectives

    Our goals for today include a battery box, repair and improvement of our intake system, and organization.

    Today's Meet Log

    • Intake redesign
    • On the robot, we are resizing the intake system as a whole so that it folds in completely and fits within the 18 by 18 sizing requirement. Our biggest focus today was on our intake system, notably building a system that deposits the minerals. We plan to create the system out of polycarb, but first we are prototyping with cardboard. There are two versions we have prototyped so far, as you can see below.

      Version 1: too wide and the triangle flaps were improperly cut so the edges interfere with the intake

      Version 2: fixes problems above, with the hypotenuse of the triangular flaps on the main part of the carrier
    • Tournament organization
    • Ethan made a battery box out of an orange REV starter kit and sawed some wood to fit snuggly in order to have some dividers. Finally we drilled a hole in the side for the power strip cord.
    • Neural network training
    • Arjun is working on our neural network for which we need to capture more training data. He is creating a program that will have the robot take pictures & capture the data we need as it drives. We had a bare-bones autonomous for the qualifier, so over the break we want to revamp our autonomous so that we can incorporate the neural network we are training more effectively. To see more about our vision training, see (E-28, CNN Training Program).

      Today's Member Work Log

      Team MembersTaskStart TimeDuration
      ArjunNeural network data collection2:001
      JanaviPrototyping2:001
      BhanaviyaPrototyping2:001
      EthanBlog2:001
      EvanBuild & Prototyping2:001
      CharlotteBlog2:001
      AbhiPrototyping2:001

    Creating Side-Latches

    Creating Side-Latches By Evan

    Task: Allow the robot's arms to stand on their own

    The issue with the lift is that many of the pieces that need to be made require specificity that’s hard to obtain using aluminum parts, so we chose polycarb. The key to making these specialized parts is a small butane torch held at just the right distance. Run the torch back and forth across the part where you want to bend, in the pattern of the bend you’re trying to achieve, watching closely for the first air pocket. Once you’ve spotted it, bend it. For tight, right angle bends, press the piece of polycarb against a hard surface until the right angle is achieved. If there’s an issue in your bend, simply heat it up again. This, however, weakens the piece so try not to do it on pieces that need to bare a load. We had to bend a complex piece, and while, it doesn’t look super complex in the picture, it had very precise requirements so that everything could slide together in unison. The part we made went in between the two linear slides on the arms to the 3d printed latch we created, and worked very smoothly. While polycarb is not the best at retaining strength over distance, it’s worked well in this instance.

    Next Steps:

    The next stage for this will be to make these brackets out of steel once we have access to the forge. This will result in a new, stronger version, which will hopefully eliminate a potential point of failure in our current robot.

    Meeting Log - Dec. 19, 2018

    Meeting Log - Dec. 19, 2018 December 29, 2018 By Ethan, Evan, Janavi, Karina, Abhi, and Arjun

    Meeting Log December 29, 2018

    Hello and welcome to the Iron Reign Holiday meet. We've got a few meet objectives today, namely:

    • Autonomous
    • BigWheel Side Plates
    • PowerPoint Revisions
    • Blog Post Backlog
    • Tent Cleanup
    These aren't all super-top priorities for us, but they all need to get done. And, as we're working with a skeleton crew, we might as well.

    So, first, Janavi, Abhi, and Ethan cleaned the tent, preparing it for autonomous testing. To do so, they got some freshmen to take up their robot parts as they cleaned and organized the field. We were missing a surprising number of tiles, so we had to replace them. As well, the recent rain had weakened the wood lying underneath. We're not going to do anything to fix this right now, but we really should.

    Next, we did PowerPoint revisions. Our presentations have always run over the 15 minute time limit, and we really need to fix it. As well, we want to change our presentation order such that we start off with the weakest award (motivate) and end on a strong note. We deleted about 5 slides, added 1, updated the Townview Tournament slide, and fixed some typos. We figure that this'll cut down our time and streamline the process.

    In the meantime, Ethan updated old blog posts and fixed broken images on the blog. Some examples of this are the Superman Arm's breakage, the old shields, Friction Test, and Battery Box posts. This took a significant amount of time.

    Finally, we had to cut new shields for the robot arms. These prevent the arms from moving back downward, allowing our robot to score in the lander. Evan measured these and melted them today, and plans to cut them next practice.

    Refactoring Vision Code

    Refactoring Vision Code By Arjun

    Task: Refactor Vision Code

    Iron Reign has been working on multiple vision pipelines, including TensorFlow, OpenCV, and a home-grown Convolutional Neural Network. Until now, all our code assumed that we only used TensorFlow, and we wanted to be able to switch out vision implementations quickly. As such, we decided to abstract away the actual vision pipeline used, which allows us to be able to choose between vision implementations at runtime.

    We did this by creating a java interface, VisionProvider, seen below. We then made our TensorFlowIntegration class (our code for detecting mineral positions using TensorFlow) implement VisionProvider.

    Next, we changed our opmode to use the new VisionProvider interface. We added code to allow us to switch vision implementations using the left button on the dpad.

    Our code for VisionProvider is shown below.

    1
    2
    3
    4
    5
    6
    public interface VisionProvider {
        public void initializeVision(HardwareMap hardwareMap, Telemetry telemetry);
        public void shutdownVision();
        public GoldPos detect();
    }
    ```
    

    These methods are implemented in the integration classes.
    Our new code for TensorflowIntegration is shown below:

     1
     2
     3
     4
     5
     6
     7
     8
     9
    10
    11
    12
    13
    14
    15
    16
    17
    18
    19
    20
    21
    22
    23
    24
    25
    26
    27
    28
    29
    30
    31
    32
    33
    34
    35
    36
    37
    38
    39
    40
    41
    42
    43
    44
    45
    46
    47
    48
    49
    50
    51
    52
    53
    54
    55
    56
    57
    58
    59
    60
    61
    62
    63
    64
    65
    66
    67
    68
    69
    70
    71
    72
    73
    74
    75
    76
    77
    78
    79
    80
    81
    82
    83
    84
    85
    86
    87
    88
    89
    90
    91
    92
    93
    94
    95
    96
    97
    98
    99
    public class TensorflowIntegration implements VisionProvider {
        private static final String TFOD_MODEL_ASSET = "RoverRuckus.tflite";
        private static final String LABEL_GOLD_MINERAL = "Gold Mineral";
        private static final String LABEL_SILVER_MINERAL = "Silver Mineral";
    
        private List<Recognition> cacheRecognitions = null;
      
        /**
         * {@link #vuforia} is the variable we will use to store our instance of the Vuforia
         * localization engine.
         */
        private VuforiaLocalizer vuforia;
        /**
         * {@link #tfod} is the variable we will use to store our instance of the Tensor Flow Object
         * Detection engine.
         */
        public TFObjectDetector tfod;
    
        /**
         * Initialize the Vuforia localization engine.
         */
        public void initVuforia() {
            /*
             * Configure Vuforia by creating a Parameter object, and passing it to the Vuforia engine.
             */
            VuforiaLocalizer.Parameters parameters = new VuforiaLocalizer.Parameters();
            parameters.vuforiaLicenseKey = RC.VUFORIA_LICENSE_KEY;
            ;
            parameters.cameraDirection = CameraDirection.FRONT;
            //  Instantiate the Vuforia engine
            vuforia = ClassFactory.getInstance().createVuforia(parameters);
        }
    
        /**
         * Initialize the Tensor Flow Object Detection engine.
         */
        private void initTfod(HardwareMap hardwareMap) {
            int tfodMonitorViewId = hardwareMap.appContext.getResources().getIdentifier(
                    "tfodMonitorViewId", "id", hardwareMap.appContext.getPackageName());
            TFObjectDetector.Parameters tfodParameters = new TFObjectDetector.Parameters(tfodMonitorViewId);
            tfod = ClassFactory.getInstance().createTFObjectDetector(tfodParameters, vuforia);
            tfod.loadModelFromAsset(TFOD_MODEL_ASSET, LABEL_GOLD_MINERAL, LABEL_SILVER_MINERAL);
        }
    
        @Override
        public void initializeVision(HardwareMap hardwareMap, Telemetry telemetry) {
            initVuforia();
    
            if (ClassFactory.getInstance().canCreateTFObjectDetector()) {
                initTfod(hardwareMap);
            } else {
                telemetry.addData("Sorry!", "This device is not compatible with TFOD");
            }
    
            if (tfod != null) {
                tfod.activate();
            }
        }
    
        @Override
        public void shutdownVision() {
            if (tfod != null) {
                tfod.shutdown();
            }
        }
    
        @Override
        public GoldPos detect() {
            List<Recognition> updatedRecognitions = tfod.getUpdatedRecognitions();
            if (updatedRecognitions != null) {
                cacheRecognitions = updatedRecognitions;
            }
            if (cacheRecognitions.size() == 3) {
                int goldMineralX = -1;
                int silverMineral1X = -1;
                int silverMineral2X = -1;
                for (Recognition recognition : cacheRecognitions) {
                    if (recognition.getLabel().equals(LABEL_GOLD_MINERAL)) {
                        goldMineralX = (int) recognition.getLeft();
                    } else if (silverMineral1X == -1) {
                        silverMineral1X = (int) recognition.getLeft();
                    } else {
                        silverMineral2X = (int) recognition.getLeft();
                    }
                }
                if (goldMineralX != -1 && silverMineral1X != -1 && silverMineral2X != -1)
                    if (goldMineralX < silverMineral1X && goldMineralX < silverMineral2X) {
                        return GoldPos.LEFT;
                    } else if (goldMineralX > silverMineral1X && goldMineralX > silverMineral2X) {
                        return GoldPos.RIGHT;
                    } else {
                        return GoldPos.MIDDLE;
                    }
            }
            return GoldPos.NONE_FOUND;
    
        }
    
    }
    

    Next Steps

    We need to implement detection using OpenCV, and make our class conform to VisionProvider, so that we can easily swap it out for TensorflowIntegration.

    We also need to do the same using our Convolutional Neural Network.

    Finally, it might be beneficial to have a dummy implementation that always “detects” the gold as being in the middle, so that if we know that all our vision implementations are failing, we can use this dummy one to prevent our autonomous from failing.

    OpenCV Support

    OpenCV Support By Arjun

    Task: Add OpenCV support to vision pipeline

    We recently refactored our vision code to allow us to easily swap out vision implementations. We had already implemented TensorFlow, but we hadn't implemented code for using OpenCV instead of TensorFlow. Using the GRIP pipeline we designed earlier, we wrote a class called OpenCVIntegration, which implements VisionProvider. This new class allows us to use OpenCV instead of TensorFlow for our vision implementation.
    Our code for OpenCVIntegration is shown below.

      1
      2
      3
      4
      5
      6
      7
      8
      9
     10
     11
     12
     13
     14
     15
     16
     17
     18
     19
     20
     21
     22
     23
     24
     25
     26
     27
     28
     29
     30
     31
     32
     33
     34
     35
     36
     37
     38
     39
     40
     41
     42
     43
     44
     45
     46
     47
     48
     49
     50
     51
     52
     53
     54
     55
     56
     57
     58
     59
     60
     61
     62
     63
     64
     65
     66
     67
     68
     69
     70
     71
     72
     73
     74
     75
     76
     77
     78
     79
     80
     81
     82
     83
     84
     85
     86
     87
     88
     89
     90
     91
     92
     93
     94
     95
     96
     97
     98
     99
    100
    101
    102
    103
    104
    105
    106
    107
    108
    109
    110
    111
    112
    113
    114
    115
    116
    117
    118
    119
    120
    121
    122
    123
    124
    125
    126
    127
    128
    129
    public class OpenCVIntegration implements VisionProvider {
    
        private VuforiaLocalizer vuforia;
        private Queue<VuforiaLocalizer.CloseableFrame> q;
        private int state = -3;
        private Mat mat;
        private List<MatOfPoint> contours;
        private Point lowest;
    
        private void initVuforia() {
            VuforiaLocalizer.Parameters parameters = new VuforiaLocalizer.Parameters();
            parameters.vuforiaLicenseKey = RC.VUFORIA_LICENSE_KEY;
            parameters.cameraDirection = VuforiaLocalizer.CameraDirection.FRONT;
            vuforia = ClassFactory.getInstance().createVuforia(parameters);
        }
    
        public void initializeVision(HardwareMap hardwareMap, Telemetry telemetry) {
            initVuforia();
            q = vuforia.getFrameQueue();
            state = -2;
    
        }
    
        public void shutdownVision() {}
    
        public GoldPos detect() {
            if (state == -2) {
                if (q.isEmpty())
                    return GoldPos.HOLD_STATE;
                VuforiaLocalizer.CloseableFrame frame = q.poll();
                Image img = VisionUtils.getImageFromFrame(frame, PIXEL_FORMAT.RGB565);
                Bitmap bm = Bitmap.createBitmap(img.getWidth(), img.getHeight(), Bitmap.Config.RGB_565);
                bm.copyPixelsFromBuffer(img.getPixels());
                mat = VisionUtils.bitmapToMat(bm, CvType.CV_8UC3);
            } else if (state == -1) {
                RoverRuckusGripPipeline pipeline = new RoverRuckusGripPipeline();
                pipeline.process(mat);
                contours = pipeline.filterContoursOutput();
            } else if (state == 0) {
                if (contours.size() == 0)
                    return GoldPos.NONE_FOUND;
                lowest = centroidish(contours.get(0));
            } else if (state < contours.size()) {
                Point centroid = centroidish(contours.get(state));
                if (lowest.y > centroid.y)
                    lowest = centroid;
            } else if (state == contours.size()) {
                if (lowest.x < 320d / 3)
                    return GoldPos.LEFT;
                else if (lowest.x < 640d / 3)
                    return GoldPos.MIDDLE;
                else
                    return GoldPos.RIGHT;
            } else {
                return GoldPos.ERROR2;
            }
            state++;
            return GoldPos.HOLD_STATE;
        }
    
        private static Point centroidish(MatOfPoint matOfPoint) {
            Rect br = Imgproc.boundingRect(matOfPoint);
            return new Point(br.x + br.width/2,br.y + br.height/2);
        }
    }public class OpenCVIntegration implements VisionProvider {
    
        private VuforiaLocalizer vuforia;
        private Queue<VuforiaLocalizer.CloseableFrame> q;
        private int state = -3;
        private Mat mat;
        private List<MatOfPoint> contours;
        private Point lowest;
    
        private void initVuforia() {
            VuforiaLocalizer.Parameters parameters = new VuforiaLocalizer.Parameters();
            parameters.vuforiaLicenseKey = RC.VUFORIA_LICENSE_KEY;
            parameters.cameraDirection = VuforiaLocalizer.CameraDirection.FRONT;
            vuforia = ClassFactory.getInstance().createVuforia(parameters);
        }
    
        public void initializeVision(HardwareMap hardwareMap, Telemetry telemetry) {
            initVuforia();
            q = vuforia.getFrameQueue();
            state = -2;
    
        }
    
        public void shutdownVision() {}
    
        public GoldPos detect() {
            if (state == -2) {
                if (q.isEmpty())
                    return GoldPos.HOLD_STATE;
                VuforiaLocalizer.CloseableFrame frame = q.poll();
                Image img = VisionUtils.getImageFromFrame(frame, PIXEL_FORMAT.RGB565);
                Bitmap bm = Bitmap.createBitmap(img.getWidth(), img.getHeight(), Bitmap.Config.RGB_565);
                bm.copyPixelsFromBuffer(img.getPixels());
                mat = VisionUtils.bitmapToMat(bm, CvType.CV_8UC3);
            } else if (state == -1) {
                RoverRuckusGripPipeline pipeline = new RoverRuckusGripPipeline();
                pipeline.process(mat);
                contours = pipeline.filterContoursOutput();
            } else if (state == 0) {
                if (contours.size() == 0)
                    return GoldPos.NONE_FOUND;
                lowest = centroidish(contours.get(0));
            } else if (state < contours.size()) {
                Point centroid = centroidish(contours.get(state));
                if (lowest.y > centroid.y)
                    lowest = centroid;
            } else if (state == contours.size()) {
                if (lowest.x < 320d / 3)
                    return GoldPos.LEFT;
                else if (lowest.x < 640d / 3)
                    return GoldPos.MIDDLE;
                else
                    return GoldPos.RIGHT;
            } else {
                return GoldPos.ERROR2;
            }
            state++;
            return GoldPos.HOLD_STATE;
        }
    
        private static Point centroidish(MatOfPoint matOfPoint) {
            Rect br = Imgproc.boundingRect(matOfPoint);
            return new Point(br.x + br.width/2,br.y + br.height/2);
        }
    }
    

    Teammember Statistics Update

    Teammember Statistics Update By Ethan

    Task: Look at the commitment changes over time of our team

    It's a new year! And, with this new year comes new tournaments, new experiences, new projects, and more. But, to grow, one must reflect. Iron Reign's had a pretty big year, from going to Worlds to the prospect of a new MXP. And, while we can't analyze every possible aspect of the team, we can look at our stats page and differences from last year to this year.

    We aren't amazing at keeping an archive of our team hours and such, so I had to pull these statistics from archive.org. The first archived version of the page in 2018 was from Feb. 14.

    As of today, our stats page displays this.

    Finally, the statistics page at the beginning of the season looked like this.

    And, the differences between each are below.

    Next Steps

    Iron Reign wishes y'all a Happy New Year! We wish to see progress among us all in these coming months.

    Off-Schedule Meeting Log, Winter Edition

    Off-Schedule Meeting Log, Winter Edition January 03, 2019 By Ethan, Evan, Karina, Abhi, and Arjun

    Meeting Log (Week of)January 03, 2019

    We have quite a few tasks this week, including:

    • Latch design

    • We've had an idea for a latch for a while. We started with the simple hook pictured below, but it was just that, a start. We want to move on to bigger and better things. So, we designed a new version, displayed below the hook.

      This version uses two of the above gears to form the latch. Then, as the robot shifts, the latch becomes undone, allowing the robot to gently fall upon the ground.
    • Latch attachment

    • So, just having a design isn't enough, it actually has to be implemented. So, Evan cut some attachment points that also function as linear slide stoppers as detailed in our last post.

      Then, we attached the latching system to the attachment posts on each side, mounting the latching system as seen here.

    • Fixing superman and wheels

    • While Karina was testing our robot, BigWheel suddenly began to lose friction, stranding itself in the middle of the field. It would only operate if more weight was put upon it. We haven't determined the reason yet; it could be that the temperature caused some strange material effect, but the new linear slides could also have shifted the weight distribution of the robot away from the main wheels. In addition, the Superman arm failed to work. We've narrowed it down to a code issue, but beyond that, we're scratching our heads.


      Karina putting weight on the robot

    • End\Beginning of year review

    • Iron Reign has a tradition of reviewing the performance of the past year; this year I chose to begin it using numbers. I went back in the archives and used the stats page to count contributions from team members. This post can be found here.
    • TensorFlow & OpenCV testing

    • We still need to fully implement gold/silver particle detection, as well as the rest of our autonomous. To begin on this long, arduous process, Abhi and Arjun worked from home to begin vision integration. At the current point, the program detects gold most of the time. We are experiencing a bug in that the telemetry isn't detected.

    Debug OpenCV Errors

    Debug OpenCV Errors By Arjun

    Task: Use black magic to fix errors in our code

    We implemented OpenCV support in our code, but we hadn’t tested it until now. Upon testing, we realized it didn't work.

    The first problem we found was that Vuforia wasn’t reading in our frames. The queue which holds Vuforia frames was always empty. After making lots of small changes, we realized that this was due to not initializing our Vuforia correctly. After fixing this, we got a new error.

    The error message changed, meaning that we fixed one problem, but there was another problem hiding behind it. The new error we found was that our code was unable to access the native OpenCV libraries, namely it could not link to libopencv_java320.so. Unfortunately, we could not debug this any further.

    Next Steps

    We need to continue debugging this problem and find the root cause of it.

    Modeling Slide Barriers

    Modeling Slide Barriers By Kenna

    Task: Create barriers to prevent the linear slide from falling

    Recently, we added polycarb barriers to our linear slide system. They were created as a temporary measure by bending the polycarb with a blow torch and are less exact than we would like.

    I originally tried to overlap 3 rectangles, and Creo didn't register it as an enclosed shape and wouldn't extrude. For any geometric shapes you want to extrude, constructed lines in sketch mode make it much easier. They ensure that everything is perpendicular, but also that your shape will still enclose so you can extrude it. Armed with constructed lines, Our models printed in roughly 30 minutes using nylon on a Taz printer.

    Next Steps

    Though nylon has its many uses, it's still not as strong as polycarb. We're looking into whether or not the printed version will withstand the slide system. Perhaps, we may need to pursue a different material or a more exact method of creating polycarb barriers. Any posts continuing this thread will be linked here.

    Meeting Log

    Meeting Log By Bhanaviya, Charlotte, Kenna, Evan, Arjun, Ethan, Janavi, Karina, Austin, Lin, Jayesh, and Omar
    Meeting Log January 05, 2019 Today's Meet Objectives

    Today's goals include lowering the latch on Superman so that it becomes more hook-friendly, taking a team picture, and re-assigning presentation slides.

    Today's Meet Log

    • Fix latch system
    • On the robot, Evan lowered the latch system so that the system would be compatible for the hanging task. After the latch system was lowered, bolts on both sides of the lift system had to be moved so that they would align with one another. See more latch updates at (E-82, Latch Model).
    • Add vision functionality to autonomous
    • In terms of code, Arjun is working on using internal Tensorflow Object Detection code to grab frames for the autonomous to avoid any bugs in the custom code he has written so far. Additionally, he is working on ensuring accuracy in the output of the OpenCV pipeline so that it will consistently sample correctly.
    • Presentation feedback from judges
    • With the alums as our judges, we did a thorough presentation run-through. A critique that persisted from our "judges" was that we weren't as enthusiastic as we could have been. So, we decided that a better way to convey our energy was by finding out a way in which we stood out from other competing teams. One way for us to stand out was the back-and-forth debate between Karina and Evan on Mini Mech vs Big Wheel. Since that interaction effectively conveyed both the iterative nature of Iron Reign's engineering process as well as our team's quirks as a whole. In the future we are going to do many run-throughs to make the presentation informative and crisp.
    • Team picture
    • And last but not least, we took a suitable team picture for our journal - this one encompassing both current and old members of the team.

    Latch Model

    Latch Model By Abhi and Justin

    Task: Model and print the Latch

    Early in the season, we made a hook, Although it was durable, it required a higher amount of precision than we would have liked to have, especially in the rushed last seconds of the endgame. As a result, we designed a latch that is completely 3D printed and placed it on the robot.

    This is the general model of it fit together (excluding left panel). The panels in the front are there to guide the latch into place when extending upwards from the ground.

    This wheel represents what actually does the latching. When sliding upwards, there are two wheels that twirl in opposite directions and slot into the lander bracket. We attached a smaller piece to this to tension with a rubber band allowing us to move up into the bracket but not back down.

    Next Steps

    We actually mounted this onto the robot and it seems to hold its own weight. However, the mounting was done very weirdly so we need to find a definite place for this system before we use it in auto or end game.

    Vision Summary

    Vision Summary By Arjun and Abhi

    Task: Reflect on our vision development

    One of our priorities this season was our autonomous, as a perfect autonomous could score us a considerable amount of points. A large portion of these points come from sampling, so that was one of our main focuses within autonomous. Throughout the season, we developed a few different approaches to sampling.

    Early on in the season, we began experimenting with using a Convolutional Neural Network to detect the location of the gold mineral. A Convolutional Neural Network, or CNN, is a machine learning algorithm that uses multiple layers which "vote" on what the output should be based on the output of previous layers. We developed a tool to label training images for use in training a CNN, publicly available at https://github.com/arjvik/MineralLabler. We then began training a CNN with the training data we labeled. However, our CNN was unable to reach a high accuracy level, despite us spending lots of time tuning it. A large part of this came to our lack of training data. We haven't given up on it, though, and we hope to improve this approach in the coming weeks.

    We then turned to other alternatives. At this time, the built-in TensorFlow Object Detection code was released in the FTC SDK. We tried out TensorFlow, but we were unable to use it reliably. Our testing revealed that the detection provided by TensorFlow was not always able to detect the location of the gold mineral. We attempted to modify some of the parameters, however, since only the trained model was provided to us by FIRST, we were unable to increase its accuracy. We are currently looking to see if we can detect the sampling order even if we only detect some of the sampling minerals. We still have code to use TensorFlow on our robot, but it is only one of a few different vision backends available for selection during runtime.

    Another alternative vision framework we tried was OpenCV. OpenCV is a collection of vision processing algorithms which can be combined to form powerful pipelines. OpenCV pipelines perform sequential transformations on their input image, until it ends up in a desired form, such as a set of contours or boundaries of all minerals detected in the image. We developed an OpenCV pipeline to find the center of the gold mineral given an image of the sampling order. To create our pipeline, we used a tool called GRIP, which allows us to visualize and tune our pipeline. However, since we have found that bad lighting conditions greatly influence the quality of detection, we hope to add LED lights to the top of our phone mount so we can get consistent lighting on the field, hopefully further increasing our performance in dark field conditions.

    Since we wanted to be able to switch easily between these vision backends, we decided to write a modular framework which allows us to swap out vision implementations with ease. As such, we are now able to choose which vision backend we would like to use during the match, with just a single button press. Because of this, we can also work in parallel on all of the vision backends.

    Another abstraction we made was the ability to switch between different viewpoints, or cameras. This allows us to decide at runtime which viewpoint we wish to use, either the front/back camera of the phone, or external webcam. Of course, while there is no good reason to change this during competition (hopefully by then the placement of the phone and webcam on the robot will be finalized), it is extremely useful during the development of the robot, because we don't have everything about our robot finalized.

      Summary of what we have done:
    • Designed a convolutional neural network to perform sampling.
    • Tested out the provided TensorFlow model for sampling.
    • Developed an OpenCV pipeline to perform sampling.
    • Created a framework to switch between different Vision Providers at runtime.
    • Created a framework to switch between different camera viewpoints at runtime.

    Next Steps

    We would like to continue improving on and testing our vision software so that we can reliably sample during our autonomous.

    Minor Code Change

    Minor Code Change By Karina

    Task: Save Bigwheel from self destruction

    The other day, when running through Bigwheel's controls, we came across an error in the code. The motors on the elbow did not have min and max values for its range of motion, causing the gears to grind in non-optimal conditions. Needless to say, Iron Reign has gone through a few gears already. Adding stops in the code was simple enough:

    Testing the code revealed immediate success. we went through the full range of motion and no further grinding occurred.

    Next Steps

    Going forward, we will continue to debug code through drive practice.

    Meeting Log

    Meeting Log January 12, 2019 By Charlotte, Kenna, Karina, Evan, Justin, Abhi, Ethan, Arjun, and Janavi

    Meeting Log January 12, 2019

    Today's Meet Objectives

    Today our goals include presentation practice, autonomous testing and fine tuning, and build changes from the newest update of the latch to replacing our REV rails with carbon fiber tubing.

    Presentation practice

    Today's Meet Log

    • Presentation practice
    • With the competition a week away, we are practicing our presentation frequently. Last time we presented, we were a bit all over the place; we talked over each other and stuttered quite a bit. This practice is to minimize these mistakes and finish our presentation in an appropriate amount of time, so we can answer questions.
    • Latch update
    • We finished up the design and print for version 2 of the latch system, and Janavi assembled it. The 2nd version changes the stopping mechanism; the bearings are now in the mount rather than in the actual sprockets. More details on this version of the latch can be found at (E-93, Latch Updates).

      Janavi & the latch
    • Lift redesign
    • Evan and Karina worked on reattaching/realigning the belt drive for the lift. It would go off in unintended angles, the process went smoothly except for the fact that we are going to need to tighen the zip ties by replacing them frequently. See more on the belt drive at (E-87, Belt Drive).
    • Carbon fiber redesign
    • The REV rails for our intake system are quite heavy, so we are building a new intake with its old components and carbon fiber tubing instead of REV rails. Justin designed and started a print for a perpendicular mounting bracket for the carbon fiber tubes.

      Justin modelling
    • Mineral storage
    • To add to the new intake system, Evan is making a new box to store minerals out of polycarb.
    • Autonomous and vision
    • Arjun tested and fine-tuned our computer vision. This vision uses Open CV, taking inspiration from the published pipeline and Doge CV. The vision is working well, so he is integrating it into the autonomous program that Abhi created. Karina and Arjun have been working diligently to test this autonomous so that it is in working condition for the next competition.
    • Side shield design
    • Ethan began the design for side shields, which we are planning to cut out using a laser cutter that is stored in our school's engineering classroom. To see more on the design process of the side shields, see (E-87, Designing Side Shields).

    Today's Member Work Log

    Team MembersTaskStart TimeDuration
    CharlotteBlog2:004
    JanaviBuild2:004
    EthanBlog2:004
    Evan/td>Build2:004
    AbhiCode & Testing2:004
    ArjunCode & Testing2:004
    KarinaBuild & Testing2:004
    JustinModelling2:004
    KennaProofreading2:004

    Belt Drive

    Belt Drive By Evan and Karina

    Task: Install a belt lift on our robot for depositing

    The most recent addition to BigWheel has been the addition of a belt drive lift on either side of the linear slides. We chose a belt lift over a string and pulley lift because it is a much more secure, closed system, and doesn’t require stringing. For these reasons, we switched to belt drive. While more complicated to build, it requires no spool, only tension, no knots, and is super smooth in its motion. Our current design relies on the same time of belt drive used on 3D printers, something that we as a team are familiar with. The issues that come with using a belt drive lift include a more complicated setup and a more difficult time to repair in the pit, a lower ability to bear weight due to slippage of the teeth, and difficulties in tensioning.

    Next Steps

    So far the belt drive has experienced a bit of slippage, but with the intake redesign we are just about to start on, it should have a better time lifting the intake.

    Designing Side Shields

    Designing Side Shields By Ethan

    Task: Create side shields for BigWheel

    Iron Reign has access to an Epilog Mini laser-cutter through our school, so we decided to use it to create side shields to protect our robot during defensive play, display our team numbers, and prevent wire entanglement

    We created our original design in Illustrator. The canvas size was 12"x18", ensuring that our design stayed within the size limits. Then, we found the side height of our robot's wheel hubs (1.3") for later use. The original design, above, was inspired by 1960s teardrop campers.

    The Epilog Mini is a CO2 laser cutter, which means that it can cut acrylic, cardboard, and wood. We don't keep our robot at school, which meant that we had to make a test cut at school. We had a variety of issues, our first print cut way too small, about 8.5"x11" when it should've been 17"x8". Our next cut caught on fire, burning in the machine as I tried to put it out without water. Our final test was successful, producing the cutout below.

    But, when fit to the robot, issues became apparent. It was barely scraping the size limit, and while it fit over the wheel mounts, it failed to match the shape of the wheel. And, the shield grazed the ground, meaning that any rotation from the Superman arm would damage it or the arm. So, we created a second, smaller design and cut it using wood, resulting in a final design.

    Selective Intake

    Selective Intake By Evan

    Task: Design a new sorting mechanism for gold and silver particles

    The differentiation between the different shapes of minerals has been something we’ve been thinking about since day one. At the time we designed a box that allowed us to sort out blocks and balls by size, but weren't able to implement it. Our original selective intake only accepted balls so we only have to go to the one loading area, but our new design allows us to deliver both blocks and balls to their respected containers. It wasn’t implemented earlier because the robot just wasn’t tall enough. With our new belt drive, It’s possible to do.

    This we decrease the difficulty of TeleOp for the drivers by increasing the speed of deposit while decreasing accuracy needed. The selective intake also has a door built into it, which holds back the minerals until we’re ready to deposit. Gravity does the rest of the work, letting the balls fall down into a brachistochrone, and letting the cubes fall through.

    The other thing that we wanted to do was to have this process be almost completely mechanical, taking more stress off the drivers. The gate is released when a lever is pushed in and translated to a quicker motion with a pair of gears at a 2:1 ratio allowing for an easy deposit. The frame of the intake is made out of polycarb, bent with the sheet bender and cut into the correct form with the bandsaw.

    The intake also uses our ice cube design from earlier in the season. It is compliant and with its new 3D-printed supports (Ninjaflex, 20% infill), it will be much more effective than the previous intake. This time, instead of stapling the thing together, we are sewing it shut, which should hopefully negate any problems the previous version had, such as coming apart where the two sides met. The intake will be offset a little from the ice tray intake to allow for as much grabbing action as possible.

    Next Steps

    Now we must allow the drive team as much time to practice with it as possible.

    Pool Noodle Intake

    Pool Noodle Intake By Evan

    Task: Design a quick intake for the robot before competition

    The night before our final qualifier, we decided that the intake system on the robot was not up to our standards. To fix this issue, we poked some holes in a pool noodle, and put surgical tubing through it. While this was a quick and semi effective fix, it did have some problems, mostly due to the rushed nature of its construction. The tubing slid back and forth, and the noodle itself was slightly offset from the depositing box, causing it to be a little off. It could only be remedied by taping all the surgical tubing together, allowing it to grip the minerals better and allow for a smoother intake. The other big issue with this version of the intake was that the depositing mechanism was imprecise and required very accurate driver control and a little bit of luck.

    Next Steps

    This isn't a permanent solution, but we need to have something simple so that we can intake the gold and silver particles at the tournament. We plan to replace this with the actual corn on the cob intake after the competition.

    Wylie East Qualifier 2019

    Wylie East Qualifier 2019 By Ethan, Charlotte, Janavi, Evan, Abhi, Arjun, Karina, and Justin

    Task: Compete at Wylie East

    Wylie East was Iron Reign's second qualifier of the year. Having qualified at the first one, we planned to use the tournament as an opportunity to practice the presentation and driver practice, as well as check up on other teams' progress. We didn't have a working robot going in - we had found that our latch was one-time-use only the night before, we had recently swapped intakes due to weight, and our autonomous was non-existent.

    Judging

    Unlike last tournament, we had actually done presentation practice, cleaned out the judging box, and revamped the presentation. We were missing a member, but we had already reassigned their slides well in advance so that people would practice them.

    And, our practice paid off. We had pretty seamless transitions, we had a good energy that the judges enjoyed, and our robot demo went really well. We got our content across, and even better, we finished way under 15 minutes so that the judges could ask us questions (even though they didn't have many to ask).

    Later, we had one group of judges come up to greet us. They mainly asked about our robot and its various functions and design choices. Our robot wasn't there, so we had to rely on old prototypes.

    Inspection

    Our robot passed field and robot inspection with flying colors and no reprimands, probably the first time that this has ever happened for Iron Reign.

    Robot Game

    Like above, we really didn't have a perfectly working robot. But, we performed much better than past tournaments due to improvements.

    Match 1

    For the first time in Iron Reign history, we tied, 211-211. Our autonomous sampled and we parked, and we were able to latch in the endgame, so it was a pretty good match all around.

    Match 2

    We lost the next match, 134-85. Our partner's robot shut down, so we couldn't keep up with the opponent. Our auto worked though, as well as latching.

    Match 3

    We lost this match, 102-237. This time, our autonomous didn't work, as our team marker fell off and knocked us off our path.

    Match 4

    We lost, 123-139. Again, our autonomous workde fine, everything else just failed.

    Match 5

    We lost, 122-154. Everything was going smoothly, but our alliance was blown out of the water during particle scoring.

    After Judging and Awards

    We weren't picked for an alliance, so we had to wait for awards. And, we ended with three awards: 1st Connect, 2nd Innovate, and 2nd Motivate. We were ineligible for Inspire due to our prior performance, but we don't believe we would have won it - the head judge stated that this was the closest tournament to regionals that we would get, so there was plenty of tough competition.

    After the awards ceremony, we came up to the fields to help clean and talk to referees. There, we were told something that we enjoyed; one of the refs told us that Iron Reign was one of the nicest and most graciously professional teams they had dealt with this season. We really liked to hear that, and it meant a lot. Also, we were told by another observer that we needed to make what our robot did more clear in the presentation, a point that we'll expand upon in the post-mortem.

    Next Steps

    See post-mortem.

    Competition Day Code

    Competition Day Code By Abhi and Arjun

    Task: Update our code

    While at the Wylie quaiifier, we had to make many changes because our robot broke the night before.

    First thing that happened was that the belt code was added. Previously, we had relied on gravity and the polycarb locks we had on the slides but we quickly realized that the slides needed to articulate in order to preserve Superman. As a result, we added the belts into our collector class and used the encoders to power them.

    Next, we added manual overrides for all functions of our robot. Simply due to lack of time, we didn't add any presets and we focused on making the robot functional enough for competition. During competition, Karina was able to latch during endgame with purely the manual overrides.

    Finally, we did auto path tuning. We ended up using an OpenCV pipeline and we were accurately able to detect the gold mineral at all times. However, our practice field wasn't setup to the exact specifications needed so we spent the majority the day at the Wylie practice field tuning depot side auto (by the end of the day it worked almost perfectly every time.

    Next Steps

    We were lucky to have qualified early in the season we could make room for mistakes such as this. However, it will be hard to sustain this, so we must implement build freezes in the future.

    Meeting Log

    Meeting Log January 26, 2019 By Charlotte, Kenna, Ethan, Justin, Arjun, Abhi, and Bhanaviya

    Meeting Log January 26, 2019

    Today's Meet Objectives

    We are going to use our experience from last week to guide our improvement until Regionals. Today we are going to discuss what these improvements exactly entail and outline a timeline for when we need to accomplish these improvements in order to allow adequate time to dedicate to autonomous code and drivers' practice.

    Today's Meet Log

    • Robot repairs
    • There were some problems with our motors: one of the axle hubs is stripped. Though we attempted to replace the axle hubs, Iron Star and Iron Core took brought most of the tools that we need to their competition.

      Karina and the robot
    • Code updates
    • We did a lot of last minute code changes during the competition. Abhi and Arjun cleaned it up and removed legacy code. Autosetup in autonomous, autonomous that works for all sides of the lander, was ditched a long time ago as it was not reliable by the time we needed to test before competition. Now that we have some time before regionals, we are bringing autosetup back. We are taking all of the code we made from scratch during the competition and integrating it into autosetup, which we hope to have ready soon to start driving practice as soon as possible.

      Coders
    • Robot model changes
    • Justin worked on the robot model. We've made lots of changes on the robot in the past month, so besides the changes that we tested on our model, it needed a couple of updates; the upgraded deposit and reinforced Superman arm. The finished robot model for BigWheel can be found at (E-107, Bigwheel Model).

    • Blog updates
    • Ethan worked on the Wylie post and the postmortem, which can be found at (T-38, Wylie East Postmortem).

    Today's Work Log

    Team MembersTaskStart TimeDuration
    AllPlanning Meeting2:10pm.25
    CharlotteTask2:004
    KennaTask2:004
    JanaviTask2:004
    EthanTask2:004
    AbhiTask2:004
    ArjunTask2:004
    JustinTask2:004

    Wylie East Postmortem

    Wylie East Postmortem By Ethan, Janavi, Charlotte, Karina, Abhi, Justin, Kenna, Arjun, and Bhanaviya

    Task: Analyze what went wrong at Wylie

    We performed well at Wylie, comparatively speaking. But, there's always room for improvement.

    Problems:

    The Robot & Code

    • Latch
    • So our first major issue was the latch. We had put together the latch the week before the tournament and tested it the night before, finding that the bearings fell out of the nylon backing under the pressure of lifting the robot. Effectively, this meant that we could only use the latch once per match as we had to reset the bearings after each use. So, we're pursuing two avenues to fix this: cut the latch in aluminum or completely redesign the latch.
    • Presets & Limits
    • Another issue that occurred was that the robot kept on injuring itself. It repeatedly overextended the Superman arm, causing its gears to disengage and strip. The same happened for the intake elbow - we didn't have limits set in so it would move too much and break. And with the belt arms, we stretched the belts out because we didn't have limits created.
    • TeleOp Helplessness
    • Another issue was that our robot didn't function well between autonomous and endgame. Our intake was recently created, and as a result, we felt it better to not attempt to score minerals. We're working on a new intake for this, some combination of our old corn-cob intake but without the Tetrix pieces that made it so heavy.

    Preparation, Presenation & Judging

    • Prep
    • We didn't pack for this tournament and as a result, we didn't have any nuts or bolts, a pretty big oversight. From now on we plan to set up boxes to bring for the week before.
    • Practice
    • Our presentation was better this time. Still, we didn't get enough practice. There were a few long pauses between people and we skipped a couple of slides. The only way to fix lack of practice is more practice.
    • Energy
    • We always need energy, it's what draws people in and gets judges to remember our presentation. Currently, we do a mini-debate within the presentation over our design choices but we plan to improve this and make it more point-by-point. In the same vein, we need to be louder.

    Pit Conduct & Misc

    • Prep Scouting Sheet
    • We need to make a scouting sheet for the tournament ahead of time with past performance. As well, we need to make a second sheet prefilled with team names for the day of. This would just reduce the amount of time spent to prepare at the tournament and transfer it to weeks-before prep.
    • Focus in pits
    • A consistent issue for Iron Reign is focus. People'll do their homework and things that aren't necessarily related to robotics in the pits and we need to stop; it always looks better to be focused when judges come around. We're still thinking of ideas to stop this.

    Latch Updates

    Latch Updates By Justin, Abhi, and Ben

    Task: Update the latch

    Our first attempt at a latch was made out of flat metal L brackets that would slide into the hook, but they slid off under any stress. We decided to make a latch with a ratchet and sprocket system. The easiest way to accomplish this was to 3d print it. There are two sprockets and the lander hook will slide in between them. This causes the sprockets to rotate and then lock, allowing the latch to support the weight of the robot. To disengage, the driver just needs to move the ratchet up and over the hook. The picture of the model shows our change in design because the right sprocket is mounted to a bearing in mount, while the left side has the bearing in the sprocket.

    The purpose of our new latch is to increase the speed of latching. The latch requires one direction of motion to fully engage it, making it perfect for autonomous. The latch also has room for error because the funnel shape of the fron plates guides the hook into the sprockets.

    Issues

  • bearings pop out under stress(fixed by moving bearings from sprocket mount to sprockets)
  • whole subsystem bends under stress(fixed by mounting the latch to aluminum instead of polycarb)
  • difficulty turning ratchet(fixed by trimming pieces)
  • Still not strong enough to support weight of robot in a match
  • Hard to get close enough to lander to engage ratchet
  • Next Steps

    We need to either strengthen our current design or find a better alternative.

    Road to Regionals

    Road to Regionals By Ethan

    Task: Consider what needs to be done before regionals

    Engineering Notebook:

    • Fix old posts, add calculations and reasons why
      • Intake posts
    • Backdate prototypes
      • Latch System
      • Superman - how we figure out what height to raise it
      • Which wheels we used based on friction
      • Which motors and why ( gear ratios and speed)
      • Gear ratio of superman
      • Linear slide vs new slides - how they work differently
      • Belt system
      •  
    • Fix timeline
    • Update posters
    • Write posts about last minute things
      • Belts
      • Autonomous
      • Latch = bad
      • Tournament
      • Post mortem
    • Create a research-like poster with all of iron reigns calculations on it
    • Create a robot manual using 3D model renders
      • Torque values, what it does, all that stuff

    Build:

    • Aluminum latch
      • Create 3D model
      • Cut at makerspace
    • Intake redesign
      • Mount red intake onto carbon fiber
      • Attach to robot
    • Front “block”
      • Create 3D model
      • Machine out of aluminum
    • Side shields
      • Fix some design problems
        • Remove points
        • Add flourishes
      • Recut with thicker acrylic
      • Mount LEDs underneath
    • Update 3D model
      • Add motors, gears
      • Update intake

    Code:

    • Auto path for crater side
      • Vision after path is complete
    • Auto path for depot side double-sample
      • Vision after path is complete
    • Auto path for crater side double-sample
      • Vision after path is complete
    • Find presets for Superman and elbow
    • Endgame mode creation
      • openCV detection of latch
      • Auto-latching and delatching
    • Autoscore during teleop

    Other:

    • Driver practice
    • Make project management charts accessible
    • Print posters
    • Make banner
    • Print banner
    • Ensure we have tent parts
    • Tent design
      • Check amount of space
      • Trophy display
      • Fairy lights
      • Organization + tool cart
    • Hats
    • Scouting Sheet
    • Tokens
      • Create design in illustrator
      • Test design
      • Cut many
    • Are we bringing things to handout?
      • Tokens to hand out ( laser cut)
      • Business Cards

     

    Superman Calculations

    Superman Calculations By Ethan

    Task: Calculate torque and other values of the Superman arm on our robot

    We want to have our robot completely replicable through the journal. So, we found it necessary to include the power calculations of various subsystems on our robot.

    Superman Arm

    The Superman arm uses two REV Core Hex motors to lift the robot upward, outputting a base 125 RPM and 6.4 Newton-meters of torque. Then, we have 15-tooth gears attached to the motors, which in turn connect to 125-tooth gears for a gear ratio of 10.4:1. Using the torque calculation WheelT=MotorT*(output/input), we find that the total torque exerted downward by the arm is 66.6 N*m.

    Then, given that the arm is .304 meters long, the upwards force produced by the Superman arm is 20.29968 Newtons. The robot itself weighs about 20 pounds, or 89 Newtons. But, since the robot is moving around its center axis, we can neglect the lower half of the robot that touches the ground with the wheels, reducing our load to 44.5 Newtons. Then, taking the integral of force with respect to the radius measured from the Superman arm, we integrate the equation force=(force at top/radius to top)*radius=292.763r. Using the limits defined by the distance to the edge of the robot (0 to .152 meters), the downward torque created by gravity is 3.38 N*m. Modeling the robot as a single point, we get this diagram.

    But, the robot doesn't always operate at optimal load. For example, when the robot is at maximum extension, there are about 60N of load above the center arm and the center arm itself is extended 18 inches, or .4572 meters. Performing the same integral as before with the new limits (0, .4572+.152=.6092), we find that the maximum possible downward torque exerted on the arm is 54.33 N*m, resulting in a net torque of 12.7 N*m upward. Superman can still raise the robot upward, but much slower and with a much greater probability of gear slippage. At these torque levels, the plastic teeth of the gears slip if they're not perfectly aligned.

    Given that the gears are composed of Acetal (Delrin/POM), that the area of one tooth is (.00104 meters * .011 meters = 0.00001144 m^2), that the arm produces 66.6 N*m * .152 m = 10.12 Newtons of force, and the Delrin/POM deformation chart, we can find that the pressure on *one* tooth of the gear is P=F/A=10.12/.00001144=884615.38 Pascals or .88461538 MPa. And, consulting the Delrin/POM deformation chart below, using the long-term line for an hour of use, we retrieve a stress of ~.5%, meaning that the teeth of the gears deform by .5% per hour of use. This alone explains our gear slippage under high loads; as the pressure on a tooth increases, they cause more deformation, which in turn results less area contact between the teeth of the gears, which results in more stress, causing a negative feedback loop.

    However, this alone doesn't explain the stripping of the gears - the gears would only deform by .0572 uM; more analysis is required. When we inspected the superman gears more closely, we found that the gears barely interlocked - maybe 1% of the gears were touching. When we go back to the pressure equation, we find that this increases the pressure on each tooth to 88 MPa. Under the short-term compression curve below, we find the strain is about 5%, or 10x the strain. This results in a deformation of about 5uM, but the contact area itself is only 104 uM, so under these loads it causes an appreciable effect.

    This leads us to the natural conclusion to solve this issue - the gears must be held tighter to increase the contact area and decrease stripping. To do this, we're starting to design a gear holder, which will be detailed soon.

    Gearkeepers

    Gearkeepers By Jose and Evan

    Task: Create and install gearkeepers to reduce slippage

    We need to install gearkeepers on the Superman arm to prevent gear skippage which damages gears over time. We designed a simple rectangle in PTC Creo and cut holes to fit bearings, 3D-printed them, and attached them.

    Now it was time to test for gear skippage. Unfortunately, we had one or two gear skips with every attempt of rotating the wheel mount. We tried using string to see if tensioning the gear holders would work but that also failed.

    We went back to the drawing board and checked for a sizing error. To calculate this we take the module of the gear and multiply it by the amount of teeth the gear has, then dividing by two to get the gear's radius. We do this for both gears and add them together. The module of the REV plastic gears is 0.75. This resulted to be (15×0.75/2)+(125×0.75/2) or 52.5 mm. And the original gear holders were 53 mm long, a slight error but at least we found the reason for error. We also noticed that there was some give in the plastic inserts for the REV bearings so we decided to tighten it down to 52mm.

    We changed the length of the inside of the gear holders from 53mm to 52mm and 3D printed them. This resulted in a complete fit where the gears were firmly engaged.

    Next Steps

    This is good for now but in the future, we need to watch the nylon of the gearkeepers for wear and tear as well as stretching - even a millimeter will allow the gears to slip.

    Intake Omnis

    Intake Omnis By Ben

    Task: Add omnidirectional wheels to intake arm

    We need to add omniwheels to the intake arm to allow the arm to rest on the ground, while still maintaining the necessary height for collecting the minerals. If the height is too low, the minerals wouldn't be able to move through the intake. If the intake was too high, it wouldn't be able to grip onto the minerals and pull them through. We decided to use omnidirectional wheels as they would allow us to drive forward and backwards with the arm extended.Our first challenge was finding space on the intake arm to attach the wheels. We had a few options:

    • Attach the wheels parallel to the arm
    • To do this, we would have to have a "u" shaped component, which we could mount off of a threaded extrusion, then attach that to the servo mount.
    • Mount the wheel perpendicular to the arm
    • This would give the same degree of maneuverability. To attach this, we would have to use an elbow bracket and attach that to an extrusion at a 90° angle.
    Both of these present a similar challenge, leaving enough room for the intake to function properly. With about 2.5in. to work with, we mounted the wheel perpendicular with a 1.75 in. extrusion. We threaded the extrusion and used an elbow bracket to mount the wheel; this ensures the strength of the wheel. This left about 0.5in. between the wheel and the "corn on the cob" intake.

    Image of wheel attached to intake arm

    Next Steps

    Our next steps are to perform testing on the wheels to determine if they are durable and low enough, and improve the performance of the robot. One issue that may arise is rubbing against the gears, as they may shift over prolonged usage, along with twisting of the extrusion.

    Mechanical Depositing

    Mechanical Depositing By Evan

    Task: Create a mechanical deposit for our selective deposit

    To relieve driver stress, we decided to put a mechanical release mechanism that would drop the minerals into the passive sorter to then further deposit them. The lever that activated the release mechanism was made of thick wire attached to a small gearbox that reversed the direction of rotation for the release gate. The lever activated the gearbox when it was pushed into the side of the lander. This created some issues that ultimately killed the mechanical release, such as a balance of tension that would never work out. We had to balance the tension of the rubber bands with the weight of the minerals while also accounting for the fact that the lever had to be pushed without our entire passive sorter being pushed beyond 180 degrees up and down.

    Because of the difficulty in implementing this, we instead switched over to a servo which now powers the release gate. While short-lived, it was a good test of the limits of our intake system, and we will be improving on it in the coming days.

    Next Steps

    We need to attach a servo to the intake with a correct mounting position to allow at-will depositing. We plan to do this with a inward-mounted servo which will then be connected to the REV hub through a wire protector, allowing us to place a servo high on the robot without worrying about the wires getting stuck in the gears and cut like before.

    Latch V.3.5 Assembly

    Latch V.3.5 Assembly By Ben

    Task: Assemble the V.3.5 latch and attach to the robot

    We assembled the fourth version of the latch today. Some of the improvements on this latch include using bigger bearings and thrust bearings inside. This latch is designed to be stronger and more reliable. After cleaning the parts and trimming some edges, we assembled the pieces. Upon assembly, we discovered an issue: the gears required a different amount of pressure to catch the lock. If left untreated, it could result in the robot falling off the hook. We determined the root of this problem was that the locking mechanism on the right gear was shorter than the left. To fix this, we trimmed a few millimeters off the piece that provides tension on the left gear to match that of the right gear.

    Latch attached to polycarbonate brackets.

    Next Steps

    We will need to perform various tests on the latch to determine if the height is correct, if the latch can support the robot, ease of latching and unlatching, and consistency. We plan to test our robot this Saturday at the DISD STEM Expo, which will provide an opportunity to practice latching.

    Fixing Mineral Dropper Components

    Fixing Mineral Dropper Components By Jose and Evan

    Task: Fix any issues with the mineral dropper

    At the STEM expo we saw a clear issue with the mineral dropper: it is very poorly geared and doesn't deposit minerals well. A quick look at the gear configuration revealed that the gears were attached in a poor manner such that there was a lot of gear skippage. To remedy this, we attached a gear-box to the dropper to keep the gears interlocked.

    The way the mineral dropper works is by having a wire attached to the shaft that turns the release be pushed when the robot hits the lander. The wire is attached with a portion of a gear custom cut for the job.

    We need to upgrade to a thicker wire for more reliable shaft rotations. After doing so we needed a different wire holder and we chose a REV wheel. After cutting it and drilling bigger holes to accommodate the new wire we needed to attach it all to the shaft for the mineral release.

    Next Steps

    We need to finish bending the wire and test its ability to open the mineral release when contacting the lander.

    Meeting Log

    Meeting Log February 02, 2019 By Charlotte, Kenna, Ethan, Bhanaviya, Jose, Ben, Evan, and Janavi

    Meeting Log February 02, 2019

    Bhanaviya working on the blog

    Today's Meet Objectives

    The DISD STEM Expo took place today. While incredibly rewarding, the experience was tiring, so only a few members made it back for the meeting that took place afterwards. This log will include our objectives and accomplishments from the meetings we held throughout the week after school which include build changes to the depositor, some calculations for analysis various parts of the robot, and preparation for our pit setup at regionals.

    Today's Meet Log

    • Design posters
    • To display this year's accomplishments, we plan to create posters for the pit. The research poster will include a few projects we have done this year including our friction tests, materials test, and torque/gear ratios calculations as well as calculations for the elbow, wheels, and other vital parts on our robot. We will also have outreach posters and a timeline of our robot design. Janavi has been designing these posters based on the journal entries we have made about the tests.

      Ethan and the research poster
    • Design passively-sorting deposit
    • Evan has been working on a mechanical depositor for minerals in the lander. We want to utilize a mechanical part to remove burden from the driver, who also has to worry about alignment with the lander as well as control of the arm. This also removes burden from our coders, who have many goals to accomplish before we will be ready for regionals. Once the initial depositor was built, we did some tests during the STEM Expo, as we had a field set up outside the MXP to show off our robots to all of the kids coming through the booth. The depositor, unfortunately, did not perform very well. The biggest problem stemmed from the elastics that enabled it to be entirely mechanical. If the elastics are too tight, it would not bend enough to let the minerals fall out of the little trap door. If the elastics are too loose, the trap door won't be sturdy enough to hold the minerals in before depositing. We are looking for other options now, and we are most likely going to opt for replacing the elastics with a driver-controlled servo. This will put more of a burden on the drivers unless the coders find the time to program sensors for depositing. Either way, we need more driving practice which we hope to accomplish in the next two weeks before regionals.

    Today's Work Log

    Team MembersTaskStart TimeDuration
    CharlotteTask4:004
    KennaTask4:004
    EthanTask4:004
    BhanaviyaTask4:004
    BenTask4:004
    JoseTask4:004
    EvanTask4:004
    JanaviTask4:004

    Code Updates

    Code Updates By Abhi

    Task: DISD STEM EXPO

    The picture above is a representation of our work today. After making sure all the manual drive controls were working, Karina found the positions she preferred for intake, deposit, and latch. Taking these encoder values from telemetry, we created new methods for the robot to run to those positions. As a result, the robot was very functional. We could latch onto the lander in 10 seconds (a much faster endgame than we had ever done).

    Next Steps

    The code is still a little messy so we will have to do further testing before any competition.

    Drive Testing at STEM Expo

    Drive Testing at STEM Expo By Ben and Abhi

    Task: Test robot performance at the STEM Expo to inspire younger kids and practice

    An FLL team gathered around Iron Reign’s robot

    We had the privilege of being a vendor and representing SEM at DISD's STEM Expo this weekend. Thousands of people cycled throughout our area during the day, so we had the opportunity to show off our robot to many people. Some of these people include FLL and VEX IQ teams, along with Best Buy volunteers. Our goal was to get kids excited about STEM and robotics, along with getting some robot practice in. We will be trying out the new latch, new presets, and prospective drivers.

    As soon as we started driving, we noticed a few issues. One of these being the belt drive repeatedly slipping. This may be a result of the belt loosening, the drive gear accelerating too quickly, heavy intake arm, or the preset causes the drive gear to keep operating, even when the arm is fully extended. We also struggled with keeping the intake box out of the way and prevent it from twisting around the “corn on the cob” intake. We will solve this by fastening the rubber band that was supposed to keep it in place. This; however, wasn’t our only intake problem. Once 2 minerals had been grabbed, they would usually fall out the intake box after lifting the arm. The intake box would turn vertical, making it easier for the minerals to shift out. This was especially an issue when trying to deposit the minerals, we would make several sudden movements, causing the arm to swing and minerals to fall out. A possible solution to this is adding a barrier between the floor of the intake box and the top of the box. This would allow for more freedom, as we could move faster without worry of losing minerals.

    Demonstrating intake arm for FLL kids

    Next Steps

    It will take a lot more practice to master latching and collecting, and even general driving. We will need to code better presets and either design a better collection box, or fix the existing one. Drivers will also have to be selected, which we will do by running several trials for each member and determining who is best at latching, scoring, and control.

    Latch 2.0 - Forged in Flame

    Latch 2.0 - Forged in Flame By Evan

    Task: Design a new latch for hanging

    Our latching system is too complicated to use quickly; it requires too much reliance on driver control and becomes jammed. So, we forged an iron hook to replace it. We started by taking an 8mm iron rod and placing it into the forge that we have, heating it up and bending it into shape over the course of an hour. We made a wire model for the hook, and then slowly and patiently formed the hook out of the rod. Then, to make an easy-to-drill connection point, we heated a section up until it was white hot and then used a punch to create a flat part that we then drilled into afterward.

    To create a mount, we took a length of steel and used an oxy-acetylene torch to heat up the areas we wanted to bend. Once this was done, we went about attaching the hook to the mount. We did this by finding the center of the mount, drilling it out, and pushing a bolt through it, surrounding all sides with washers. We then mounted a servo next to the hook and attached it with a piece of wire, which was secured to the hook by two notches cut out of either side of the tail of the hook. Later, after finding the wire to be too flimsy, we attached the two together with a strip of polycarb. It works well, allowing us to mount and dismount much easier than we would have hoped for with our last latch. While the last latch was purely passive and required no electrical components, this one gives us much more control in how we latch and delatch.

    Meeting Log

    Meeting Log By Charlotte, Evan, Ethan, Kenna, Karina, Abhi, Arjun, Ben, Jose, and Janavi
    Meeting Log February 09, 2019

    Today's Meet Objectives

    Today we participated at a scrimmage held at Woodrow Wilson High School. This was a fantastic opportunity to get some driver practice in real, timed games and adjust for issues.

    Today's Meet Log

    • Hook implementation
    • Since we have made a few changes to the robot, such as adding a servo to our previously mechanical output mechanism, we evaluated how well they worked. We wired the servo and fixed the wiring from the arm that got tangled in the motor using a wire router to take control of this issue. As well, we began auto tuning for the new hook.

      Fire from the forge from crafting the hook

      The burning metal being bent into our hook
    • Driver practice
    • When we weren't making changes on the robot, we were practicing driving. Some difficulties we faced included getting stuck in the crater because of our arm and the disconnection of our hook from the servo horn due to our attachment with zipties. When we got back to the house, we began changes to fix these issues by creating a replacement for the zipties out of polycarb and working on presets to improve the balance of the robot.

    Woodrow Wilson Scrimmage

    Woodrow Wilson Scrimmage By Bhanaviya, Charlotte, Janavi, Kenna, Karina, Evan, Abhi, Jose, Ben B, and Arjun

    Task: Compete at the Woodrow Wilson Scrimmage with Woodrow teams

    This past Saturday, Iron Reign competed in the Woodrow Wilson Scrimmage. To ensure that the wiring did not become tangled when the robot moved around, we added an ABS cable-carrier to the arm of the robot.

    Overall, Iron Reign was able to establish a semi-stable deposit scoring game-plan in the match. Since we haven’t focused on practicing game play in a while, this scrimmage gave us an opportunity to pin-point build and code issues, as well as get a clearer idea of what our strategy for regionals needed to look like.

    Next Steps

    We are incredibly thankful for Woodrow Wilson and their teams for hosting us, as well running such an effective scrimmage. The opportunity to connect with other teams in our region has given a clearer idea of what we can learn from the teams around us to improve our overall team presence.

    Bigwheel Model

    Bigwheel Model By Justin

    Task: Design and update the Bigwheel Model

    We are updating our bigwheel model to represent our current robot. We had a model of just the chassis from the chassis study, so we are currently adding all of the changes we made throughout the season.

    Completed Changes

  • Added current intake
  • Added sorting system
  • Modeled the linear slide lift
  • Modeled superman arm
  • Future Changes

    The lift has been changed recently so the model needs to be updated. The main problem with this is that the new slides are not standard parts, so there are no accurate CAD files. This means we have to custom model our new slides to maintain accuracy with our model. The motor placement on the chassis needs to be fixed because we the measurements were estimates. There are many small 3d printed parts that need to be added to the robot, as well as our new ratchet latch.

    Next Steps

    We need to work on future changes and get our model up to date with our robot so we can start conceptualizing new subsystems.

    BigWheel Upgrades

    BigWheel Upgrades By Evan

    Task: Fix some issues on BigWheel before the build freeze

    We made more secure way of activating our hook, so we switched our piece of wire attaching the servo to the hook with a much stronger and less likely to bend strip of polycarb, which greatly improved the reliability of the hook.

    As well, we limited the back and forth motion of our slides at their attaching points. I achieved this by inserting a small piece of drywall sandpaper in between the stages of the slides. Hopefully, the added friction will create a stronger hold between the stages fo the slide.

    Next, we ground down a bolt to more securely attach the servo horn to the servo since it’s a REV hex shaft to servo adapter and the bolts we had didn’t fit inside well enough. Once that was done, we changed the ratio between the belt drive pulleys, going from a 1:1 (36 teeth to 36 teeth) to a 5:3 (60 teeth to 36 teeth) by increasing the size of the pulley at the motor. This should increase the quickness of our lift and hopefully let us squeeze a few more mineral pick up and depositing cycles in.

    Next Steps

    It's time to turn the robot over to the coders and drivers, so there won't be many changes after this,

    Pulley Spacers

    Pulley Spacers By Ethan

    Task: Design and implement pulley spacers to prevent belt interference

    We had an issue where the belts that allowed our arm to slide upward were misaligned, resulting in the belts frequently slipping. We narrowed the slippage down to a single point, at this pulley.

    We had to create a new spacer to keep that section of the belt inline with the rest. As usual, we took measurements and replicated them in Creo. It had to be about 3.5 centimeters long, the same width of the metal plate. The depth of the indentation to attach to the linear slide is about 0.75 centimeters and the diameter of the M3 holes 3 millimeters. With these measurements, we designed the piece and printed it in 60% infill nylon, strong enough to withstand the weight of the linear slides. This is what version one looks like:

    However, this version's holes were too far down, allowing the toothed sections of the belts to interact and jam. So, we decreased the height of the bottom pulley-holes so that the middle section of the belt would slider higher up, preventing interference. This resulted in the final version seen at the top of the article.

    Next Steps

    We still have to fully test these spacers, but we can't do a full test until we fix the gears supporting the elbows, which will be detailed in another post.

    Robot Issues - Gear Grinding

    Robot Issues - Gear Grinding By Ethan

    Task: Analyze the issues with the elbow arm of our robot

    The elbow arm of the robot is what allows us to rotate the arm of our robot - the linear slides what hold the intake. Recently, while doing some drive testing, we found that the elbow wasn't acting as it should. When we took a closer look at it, we realized that the metal gears had started to destroy each other.

    Before installing new gears and just having the same thing happen again, we wanted to analyze why this was happening. Remembering that pressure is equal to force divided by area, we noticed that the gears weren't fully interlocking, reducing the area for force to act on. And, the teeth of these gears are minuscule things, so the pressure on each one is immense, even more so with the full torque of the extended linear slide behind them. And, while these aluminum gears may not bend that well, under these immense pressures, they sure can break since hardness and brittleness trade off. And, even then, with high pressure and frequent use, they can still easily grind down, resulting in this scene:

    But, that's not all. When we tried to run the elbow, we realized that the motor shaft themselves were out of alignment. This is hard to capture in a single picture, but this manifested itself as a sort of wobble when the motor was repeatedly run. With full, non-ground gears, this would probably be fine, but the moving shaft reduced the area of interaction between gears, contributing to the gear-dust all over our robot. Finally, as the gears were reduced to almost nothing, this wobble made it so the gears wouldn't interact at all.

    The solution to this is complicated, as we only have one set of spare gears. If we had more, we would be able to replace them as needed, but currently, we couldn't guarantee that they wouldn't give out at regionals. First, we need to replace the motors, as any wobbliness reduces the area of interaction between gears, which increases the pressure on the teeth accordingly. Then, we need to create gearkeepers to hold the gears to maximum contact. We've created gearkeepers before under the same circumstances for the arm that lifts our robot up (we had a similar gear-stripping scenario), but this may not be enough alone. First, we use metal gears on the elbow, which have smaller teeth area-wise than the plastic ones elsewhere. Plus, the gearkeeper design below doesn't compensate for any later wobbliness that may occur and may wear out itself, as its essentially a nylon strap between two shafts. So, we need to design a gearkeeper that doesn't only attach from shaft to shaft but shaft-shaft-robot, as this would prevent the pesky wobbliness and decrease tooth pressure as much as possible.

    Next Steps

    We've forwarded this analysis to the modeling team, who will produce a print later this week so that we can bring our robot back up to snuff.

    Research Poster

    Research Poster By Janavi and Ethan

    Task:Create a Poster amalgamating all of our math

    Throughout this season our team has completed various calculations from the torque of our robotics arm, to the speed of the wheels. Since these calculations are spread throughout our journal, we decided to amalgamate them into a single poster that is easy for us to refer to. In this poster we have calculations for

    • Torque/ Gear Ratios
      • Intake Arm Torque
        • (Robot Manual)
      • Wheel Gear Ratios and Speed Calculations
        • (E-132, Intake Speed )
        • (E-52, Linear Slide Lift)
      • Elbow Torque and Gear Ratios
        • (Robot Manual)
      • Superman Torque and Gear Ratios
        • (E-95,Superman Calculations)
      • Superman Gear Material Calculations
        • (E-95,Superman Calculations)
    • Friction Tests
      • (E-59,Friction Coefficient and Energy)
      • Coefficient of Friction of Silicone Intake
        • (E-59,Friction Coefficient and Energy)
    • Material Testing
      • Linear Deformities with Nylon
        • (E-62,Linear Nylon Strength Test)
      • Linear Deformities with ABS
      • Linear Deformities with PLA

    Elbow Rebuild

    Elbow Rebuild By Ethan, Jose, Karina, and Ben

    Task: Rebuild the elbow after total gear annihilation

    In a previous post, we detailed the extent to which we had stripped our gears - they were missing teeth in several places and the black anodization layer had completely stripped away. So, we had to replace them. The first order of action was to design gearkeepers for them. We've designed gearkeepers before, for the Superman arm, but these have different requirements. They must connect the gears on both elbow driver and slave, but also must mount to the robot itself to prevent the motor shaft from wobbling, which had previously caused major issues. We came up with this design, printing it out in 60% infill nylon.

    The next thing to do was replace the actual gears. To do so, we had to dismantle the entire elbow and replace the gears and shaft collars. This alone took about two hours per side. We added the new gears, ensuring that they were in alignment, and printed a circular part to mount the top of the gears to the linear slide so that the entire system would rotate when the gears were turned. Then, we remounted the belts and aligned them. After, we attached the new gearkeepers, ensuring that the gears interlocked perfectly.

    Next Steps

    So far, we haven't experienced issues with the new elbow, but we're getting our hands on a new set of gears to be safe. We expect this system to continue to work for the Regional tournament, and are performing drive practice to ensure this.

    Meeting Log

    Meeting Log February 16, 2019 By Ethan, Janavi, Kenna, Justin, Bhanaviya, Ben, Abhi, and Arjun

    Meeting Log February 16, 2019

    So, its the last week before Regionals, so we have a lot of work to do, from robot work to presentation.

  • Linear slide arm repairs
  • We started off the day with working on the elbow for the arm. For the past week, we've been dealing with the gears on the elbow stripping. So, we replaced the gears on both sides, threadlocked the motors so that the shafts wouldn't wobble, and installed upgraded triangular gearkeepers so that that that that the gears would fully interlock, preventing the gears stripping. This process took about 90 minutes per side, taking up time we needed for autonomous. But, our build freeze has persisted - we haven't added anything else. In the same vein, Justin worked on the 3D model, integrating the corncob into the design.

  • Blog updates
  • We're also trying to finalize our journal, so we're finishing up posts. Janavi was working on a post about the research poster; Arjun was working on computer vision posts; Abhi was updating code posts. Ethan was going through and retagging posts so that the table of contents is accurate, fixing the posters we're printing, and updating presentation photos. Janavi and Kenna were also working on the handouts for Regionals.

  • Driver practice
  • Since Karina isn't here, we're letting Ben practice driving. We're consistently getting 2-3 cycles in the lander with him as opposed to Karina's 4-5, but practice will help. He's not all there yet, he crashed the robot somehow, but its a start. We're also working on autonomous delatch and tuning as he drives using telemetry data.

    Today's Work Log

    Team MembersTaskStart TimeDuration
    AllPlanning Meeting2:10pm.25
    EthanEdit blog posts and update posters2:004
    EthanFix robot gears12:002
    Justin3D Model2:004
    BhanaviyaComputer Setup2:001
    KennaDesign handouts2:004
    JanaviBlog posts2:004
    BenReplace gears2:004
    AbhiRobot tuning2:004
    ArjunControl Award2:004

    Latch Designs - A Retrospective

    Latch Designs - A Retrospective By Ethan

    Task: Analyze past successes and failures in our latching system

    Version 1

    The first version of the latch worked decently. We started out with the idea of a one-way, passive latch. This idea involved mounting smaller bearings and gears between them, with a spring-like nylon piece that moved only when downward pressure was placed upon the gears. This design was only fully realized before the Wylie Qualifying tournament, and only tested the night before. We found that the bearings popped out under pressure necessitating a reset after every match and meaning that we could only latch once per match. We opted for the endgame latch, as it was more reliable. But, this cut the amount of points we could receive immensely. After the tournament, we decided to do a full redesign.

    Version 2

    The second version's changes were simple. We redesigned the nylon "spring" and made it thicker and more prominent. This made it so the latching gears were more firm than before, which in turn allowed more weight to be put on them. However, the issue with the gears was still present; as the load on the latch increased, the nylon would bend more and more, allowing the bearings to fall out so that the latch would jam in place. This version was quickly scrapped.

    Version 3

    At this point, we were sick of the bearings popping out. So, we widened the holes immensely to fit larger bearings which in turn had larger holes allowing for bolts to be run through. This was overkill, but it ensured that no slippage would occur during normal robot usage. Again, we also thickened the nylon "springs" so that the gears would stay in place without significant upward force.

    We realized, that while technically impressive, the latch as we knew it had to go. It worked, but it was too time-costly to justify using, as the driver had to precisely line up the bot next to the lander to use it, taking about 20 seconds. In addition, it was difficult to code as it required several intricate simultaneous robot operations: the lift needed to descend at the exact same moment Superman needed to rotate, all while the elbow rotated the robot 90 degrees. In summary, it was an overly burdensome task. So, we threw away all that work, these past two months of labor in favor of a simpler option.

    Version 4 - the Hook

    We decided that it was time to go back to the drawing board. In time periods, it was approximately a jump from the current era to the Iron Age. So, we designed appropriately. We designed a stainless steel hook, first making one out of prototyping wire. Then, we heated up the forge, adding plenty of coke, and set to work. We chose a stainless steel rod, 8mm in diameter and warmed it to red hot, beating out the initial design. We let the initial rod air cool so that it would be soft enough to drill through, creating the mounting point for the robot. Then, we reheated it to its critical point (when it loses its magnetic properties) and quickly quenched it to reharden it. But, simply quenching it makes the steel too brittle to use in competition, so we finished up the hook by tempering it, using an oxy-acetyline torch on it until the surface became matte. Finally, we had the hook seen above. After all that work, we'd gone with the simplest option because sometimes, it is the best.

    Off-Schedule Meeting Log - Week before Regionals

    Off-Schedule Meeting Log - Week before Regionals February 19, 2019 By Ethan, Evan, Jose, Charlotte, Karina, and Justin

    Meeting Log February 19, 2019

    It's the week before Regionals, so the house is a flurry of activity - all hands are on deck for every possible facet of the team.

    Monday

    The week started out with three projects. Justin worked on the robot model, taking measurements for the intake and putting the assembly together for six hours, completing the model. Just as he left, Ethan started the editorial review. The goal of the review was to develop a more cohesive journal, a journal that could easily be flipped through. The list of tasks created from this session are below. In addition to this, Ethan worked on making an LED hat for the tournament.

    Editorial Review Listing

    • Unbury $150k grant post, make title "major grant to fund replacement of vehicle" + fix receiving + remove last sentence
    • Add Ben, Jose headshots to organization slide
    • Replace townview qualifier photo
    • Add Microsoft section to stem expo post
    • Add motivate tag, remove connect from drive testing at stem expo
    • Remove all motivate from connect table of contents
    • Bold totals in the iron reign grants post
    • Fix pulley spacers image
    • Fix broken image last meetinglog
    • Remove connect posts from motivate table
    • Add to stem spark post
    • Post summary of motivate and connect 2x
    • Add letters to presentation tabs to indicate award
    • Change decisions to priorities in presentation
    • Remove center photo collector system in presentation
    • Delete slide 38 with wordcloud in presentation
    • Remove what we need help with slide
    • Make text bigger on Connect summary slide, add totals in title in red
    • Update journal summary
    • Make a latch retrospective post
    • Post about rebuilding elbow
    • Fix Woodrow Code blog post
    • Add articulation and drive enhancement posts
    • Make post about bearings in linear slide

    Tuesday

    • Battery Mount
    • Evan worked on a battery mount for the robot. While drive testing, we had found that the battery and camera would fall out under extreme conditions, so we decided to create a new one. Evan cut battery "corners" out of polycarb and mounted them together, ensuring that the battery would stay static in every match.
    • Editorial Review 2
    • Ethan wrote new posts on the history of the latch, the rebuilding of the elbow, connect and motivate summaries, and fixed the above issues from the editorial review. In addition, we rewrote the summary, as we found that it would be heavily considered in Regional judging. Charlotte uploaded old meeting logs.
    • Driver Practice
    • Karina, Justin, and Jose practiced driving the robot. We discovered that the robot latches extremely well with the new hook and that the autonomous delatch works. We also tested the articulation, or poses, of our robot. The only issue that popped up was when the robot moves into deposit mode, it tips toward the side with linear slides, but Karina discovered that if she drives the robot forward at the same time, she can ram the robot into the correct position. Karina got to 4-5 cycles per match with the new updates. This practice was a way to test the strength of our robt - we've had our robot break under stressful situations previously - and this time nothing broke. The biggest issue was that a servo wire on our intake came unplugged, but even with that, our robot still worked.
    • Model Articulation
    • Justin took the last measurements for the model of our robot, then started to take pictures of the articulations we made in the code.
    • Hat
    • We finished the light-up LED hat.

    Wednesday

    • Driver Practice and Autonomous
    • Karina and Abhi worked on the robot. Karina gave advice for improving our robot's articulations to Abhi, who proceeded to fix the code for better driver practice. Abhi also worked on delatch in autonomous, reversing the autonomous driver enhancement code and taking data from Karina's testing. We discovered one new issue with our robot, that the gearkeepers for Superman pop out of alignment after about 100 uses. All we need to do is realign them, and they'll be back to full functionality.
    • Control Award
    • Arjun continued writing the Control Award submission, adding in the new articulations and poses of the driver enhancements. Janavi created state diagrams for the code to add to the submission.

    Big Wheel Articulations

    Big Wheel Articulations By Abhi

    Task: Summary of all Big Wheel movements

    In our motion, our robot shifts multiple major subsystems (the elbow and Superman) that make it difficult to keep the robot from tipping. Therefore, through driver practice, we determined the 5 major deployment modes that would make it easier for the driver to transition from mode to mode. Each articulation is necessary to maintain the robot's center of gravity as its mode of operation shifts.

    The position seen above is called "safe drive". During normal match play, our drivers can go to this position to navigate the field quickly and with the arm out of the way.

    When the driver control period starts, we normally navigate to the crater then enter the intake position shown above. From this position, we can safely pick up minerals from the crater.

    From the intake position, the robot goes to safe drive to fix the weight balance then goes to the deposit position shown above. The arm can still extend upwards above the lander and our automatic sorter can place the minerals appropriately.

    During the end game, we enter a latchable position where our hook can easily slide into the latch. After hooked on, our robot can slightly lift itself off the ground to hook.

    At the beginning of the match, we can completely close the arm and superman to fit in sizing cube and latch on the lander.

    As you can see, there is a lot of articulations that need to work together during the course of the match. By putting this info in a state machine, we can easily toggle between articulations. Refer to our code snippets for more details.

    Next Steps

    At this point, we have 4 cycles in 1 minute 30 seconds. By adding some upgrades to the articulations using our new distance sensors, we hope to speed this up even more.

    Cart Hack

    Cart Hack By Arjun

    Task: Tweaking ftc_app to allow us to drive robots without a Driver Station phone

    As you already know, Iron Reign has a mechanized cart called Cartbot that we bring to competitions. We used the FTC control system to build it, so we could gain experience. However, this has one issue: we can only use one pair of Robot Controller and Driver Station phones at a competition, because of WiFi interference problems.

    To avoid this pitfall, we decided to tweak the ftc_app our team uses to allow us to plug in a controller directly into the Robot Controller. This cuts out the need for a Driver Station phone, which means that we can drive around Cartbot without worrying about breaking any rules.

    Another use for this tweak could be for testing, since with this new system we don't need a Driver Station when we are testing our tele-op.

    As of now this modification lives in a separate branch of our code, since we don't know how it may affect our match code. We hope to merge this later once we confirm it doesn't cause any issues.

    Up-to-Date Bigwheel Model

    Up-to-Date Bigwheel Model By Justin

    Task: Finish the Bigwheel model

    Updating the Bigwheel model to the robot’s current configuration was a challenge. The new linear slides are not standard parts, so we had to model them from scratch. There was some cleaning up that was needed on the drivetrain of the model. This was mainly attaching floating motors to motor mounts and axles to bearings. These were mainly cosmetic changes, but they help define the purpose of the different parts of the drivetrain. We also updated the intake assembly to our current ice cube tray intake. The structure of the intake was easy to model but the ice cube tray gave us some trouble with its unique shape and pattern. The ratchet latching system was a failure, so a new hook model was needed. The main issue with this was that we custom forged our new hook, so there was some difficulty in getting the model to accurately represent the capabilities of the hook. Another challenge was the mineral storage system. This is made from polycarb pieces and has many unique pieces, so arranging the pieces to accurately show the flow of minerals was difficult.

    In addition to updating the model, we also learned how to show the different movements of the robot with the model. Mechanical constraints were added to allow certain parts to slide or rotate. The one problem we had with this was that there were no limitations to how far something could slide or rotate, so many parts of the model would disconnect and be left floating. After some research, a solution was found. Zero points were created for each moving part and minimum and maximum movement limits were added. Some parts that now can move on the robot are the wheels, superman arm, hook, and linear slides. This allows us to not only show the movement of the robot, but also the limitations of its parts, which can help us visualize new solutions to our remaining problems.

    Next Steps

    Our next step is to wait for more build changes, so we can keep updating the model. Another addition we might make is making stress maps of the robot in different configurations to see where parts might fail. This has been an ongoing challenge of keeping the model accurate when the robot gets updated or rebuilt, and now we finally have a finished model and ready robot for regionals.

    Wylie Regionals 2019

    Wylie Regionals 2019 By Ethan, Charlotte, Evan, Kenna, Karina, Abhi, Arjun, Bhanaviya, Ben, Justin, Jose, and Janavi

    Task: Compete at the North Texas Regional Tournament

    Preparation

    Unlike other tournaments, we started packing before morning. We packed as if we were going out of state, bringing a bandsaw, all-new charging box, every replacement part imaginable, and a printer which would ultimately come in handy later. We relied on a packing list created by Janavi, detailed below.

    Because of this, we got to Wylie on time, turned in our notebooks, had the team rosters printed out, and were able to start right away.

    Inspection

    Breaking our all-season streak, we failed our first inspection, cited for our unruly cable management. So, we made a hasty retreat back to the pits and zip-tied the cables together and rethreaded our intake servo wires through the cable guards, then brought it back to inspection. We passed, but we were warned about possible size issues with the team marker. But, looking at RG02, we realized that it wasn't a major concern.

    Judging

    The main issue this time was not speed or knowledge but simple enthusiasm - it just felt off and a little uncharismatic. However, we received three separate pit visits for what we believer were Motivate, Connect, and Innovate. In particular, we were able to get the Motivate judges out to see the MXP and talk about expanding the program while keeping it sustainable. The Innovate judges focused on the Superman mechanism, as it's fairly unique, and we fielded questions about the design process. In Connect, we also talked about the MXP and its $150k grant largely because of our efforts.

    Robot Game

    Match 1(Q3)
    For the first time in the Rover Ruckus season, we won a game. Both us and Corem Deo had almost perfect auto and Corem Deo got plenty of mineral cycles into the lander. Unfortunately, BigWheel tipped over during end game so we couldn't latch. However it did not affect the match results significantly.

    Match 2(Q9)
    Unfortunately, we lost. Both our autos failed in some way and BigWheel ended autonomous with one wheel in the crater, wasting us 30 seconds during teleop just to get out. Also, most of our mineral cycles failed and we couldn't latch during end game and had to partially park in the crater.

    Match 3(Q15)
    To our surprise, we won. We were against Elmer and Elsie, who were seeded 1st before this match. We had a perfect auto this match while the other side had some issues with their's. During teleop we had some pretty successful mineral cycles and both robots hung onto the lander with the other side only having one hang and one robot partially parked.

    Match 4(Q26)
    We didn't expect to pull a third win but we did. Our auto also failed a little again but it didn't cost us any time during teleop like last time. We also had some very successful mineral cycles this time, but when attempting to hang BigWheel tipped when going into its preset position for hanging, even so, it didn't affect match results.

    Match 5(Q33)
    Once again we didn't expect a fourth win, but it happened. Before this match we wanted to test our autonomous with the Lamar Vikings to check if the robots would collide during autonomous, but due to mechanical issues on their side this was delayed and we had to queue without doing so. Indeed, our robots collided in the depot causing us to miss out on 75 points. During teleop one robot on the other side disconnected but on our side two of our servos disconnected, the mineral gate and the hook, so we couldn't score minerals or latch so we played some minor defense and partially parked in the crater.

    Match 6(Q36)
    Our luck ran out in this match as we lost. This was a very tight match against TechicBots, the first seed. Both sides ended autonomous 150-150. The mineral game was also tight, the lead switched between both sides many times as minerals were scored but the other side took the lead once BigWheel tipped over. We couldn't hang once again and both our opponents kept scoring, leading to our loss.

    For the first time this season, we were selected for Semis as the first pick of the third alliance.

    Match 1
    We lost. Our autonomous failed as well as teleop while the other side continuously scored minerals into the lander. And yet again we couldn't hang due to tipping.

    Match 2
    We lost again. We began a timeout due to technical issues with the phones and ultimately had to give up and leave BigWheel to sit idle on the field for two minutes and thirty seconds while the Lamar Vikings attempted to win without us.

    Awards Ceremony

    By the time the ceremony started, most of us had been up for 13+ hours, so we were all a little under the weather. We first received the Motivate award! It's always nice to have your efforts recognized and this was no exception. The Motivate award means a lot to us - it's what we got last year at Worlds. Then, we heard, "3rd place Inspire Award goes to...team 6832 Iron Reign!" And the SEM section went wild. We advanced!

    Next Steps

    The post-mortem will be in a later post. See y'all at Worlds!

    North Texas Regional Postmortem

    North Texas Regional Postmortem By Ethan, Charlotte, Abhi, Janavi, Evan, Ben, Jose, Justin, Karina, Bhanaviya, and Arjun

    Task: Analyze what went wrong at North Texas Regionals

    We performed really well at Regionals; we actually won our first game of the season and ended 4-2 and were selected for an alliance. But, we still didn't do everything right. We were on the verge of not being chosen for Inspire, and we can't risk the same at Worlds.

    Problems:

    The Robot & Code

    • Auto & Setup
    • To begin, we had issues with preparing our robot, particularly that we didn't have enough practice setting it up for autonomous. As well, we didn't have a way to verify that the setup was correct.
    • Initial Code
    • We had high pings at the tournament, so we plan to reduce our telemetry to two lines. As well, our control scheme was too complicated, and we need to simplify it.
    • TeleOp
    • The robot kept tipping because of the complicated management of three systems. When in motion, angular momentum is conserved, making it hard to manage the robot and keep it upright. As well, we couldn't see the minerals in the intake.
    • Build
    • Again, we couldn't see the minerals in the intake. As well, the carbon fiber intake rod broke along with the battery and phone mount. These all necessitate redesigns. Finally, our wiring was out of hand.

    Pit Interviews

    • MXP was not set up for Motivate judges
    • Missed groups of judges looking for our robot several times
    • Didn't let judges leave when they wanted to - kept on talking

    Pre-tournament Preparation

    • Presentation
    • We hadn't practiced the robot demos; our IMU demo worked but the latch demo didn't. As well, we hadn't done a runthrough before handing out items from our presentation box. So, more thorough presentation practice is needed.
    • Engineering Journal
    • The team as a whole needs to focus on getting their blog posts in on time. It's hard to prepare the journal when not all posts needed for it are present. As well, we forgot to print the cover sheet for the control award.

    Pit Setup & Conduct

    • Ugly Pit
    • Our signs were disorganized and not easy to view, and our pit in general was a mess. We didn't have handouts, and our activities were off-topic.
    • MXP Setup
    • Even though the MXP is a centerpiece of our presentation, we left it wrecked after we unloaded all of our materials and making it too dirty for a tour.
    • Team Members
    • A few team members were not actively participating at the tournament, giving a bad impression for the judges.

    Road to Worlds Document

    Road to Worlds Document By Ethan, Charlotte, Evan, Karina, Janavi, Jose, Ben, Justin, Arjun, and Abhi

    Task: Consider what we need to do in the coming months

    ROAD TO WORLDS - What we need to do

     

    OVERALL:

    • New social media manager (Janavi/Ben) and photographer (Ethan, Paul, and Charlotte)

     

    ENGINEERING JOURNAL: - Charlotte, Ethan, & all freshmen

     

    • Big one - freshmen get to start doing a lot more

     

    • Engineering section revamp
      • Decide on major subsystems to focus on
        • Make summary pages and guides for judges to find relevant articles
      • Code section
        • Finalize state diagram
          • Label diagram to refer to the following print out of different parts of the code
        • Create plan to print out classes
        • Monthly summaries
      • Meeting Logs
        • Include meeting planning sessions at the beginning of every log
          • Start doing planning sessions!
        • Create monthly summaries
      • Biweekly Doodle Polls
        • record of supposed attendance rather than word of mouth
      • Design and format revamping
        • Start doing actual descriptions for blog commits
        • More bullet points to be more technical
        • Award highlights [Ethan][Done]

    Page numbers [Ethan][Done]

        • Awards on indexPrintable [Ethan][Done]
      • Irrelevant/distracting content
        • Packing list
        • Need a miscellaneous section
          • content
      • Details and dimensions
        • Could you build robot with our journal?
        • CAD models
        • More technical language, it is readable but not technical currently
    • Outreach
      • More about the impact and personal connections
      • What went wrong
      • Make content more concise and make it convey our message better



    ENGINEERING TEAM:

     

    • Making a new robot - All build team (Karina & Jose over spring break)

     

      • Need to organize motors (used, etc)
      • Test harness for motors (summer project)
    • Re-do wiring -Janavi and Abhi
    • Elbow joint needs to be redone (is at a slight angle) - Justin/Ben
      • 3D print as a prototype
        • Cut out of aluminum
      • Needs to be higher up and pushed forward
      • More serviceable
        • Can’t plug in servos
    • Sorter -Evan, Karina, and Justin
      • Sorter redesign
    • Intake -Evan, Karina, Abhi, Jose
      • Take video of performance to gauge how issues are happening and how we can fix
      • Subteam to tackle intake issues
    • Superman -Evan and Ben
      • Widen superman wheel
    • Lift
      • Transfer police (1:1 to 3:4)
      • Larger drive pulley
        • Mount motors differently to make room
    • Chassis -Karina and a freshman
      • Protection for LED strips
      • Battery mount
      • Phone mount
      • Camera mount
      • New 20:1 motors
      • Idler sprocket to take up slack in chain (caused by small sprocket driving large one)
    • CAD Model



    CODE TEAM: -Abhi and Arjun

    • add an autorecover function to our robot for when it tips over
      • it happened twice and we couldn’t recover fast enough to climb
    • something in the update loop to maintain balance
      • we were supposed to do this for regionals but we forgot to do it and we faced the consequences
    • fix IMU corrections such that we can align to field wall instead of me eyeballing a parallel position
    • use distance sensors to do wall following and crater detection
    • auto paths need to be expanded such that we can avoid alliance partners and have enough flexibility to pick and choose what path needs to be followed
      • In both auto paths, can facilitate double sampling
    • Tuning with PID (tuning constants)
    • Autonomous optimization



    DRIVE TEAM:

    • Driving Logs
      • everytime there is driving practice, a driver will fill out a log that records overall record time, record time for that day, number of cycles for each run, and other helpful stats to track the progress of driving practice
    • actual driving practice lol
    • Multiple drive teams

     

    COMPETITION PREP:

    • Pit setup
      • Clean up tent and make sure we have everything to put it together
      • Activities
        • Robotics related
      • Find nuts and bolts based on the online list
    • Helping other teams
    • Posters
    • Need a handout
    • Conduct in pits - need to be focused
    • MXP or no?
    • Spring break - who is here and what can we accomplish
    • Scouting

     

    Code Refactor

    Code Refactor By Abhi and Arjun

    Task: Code cleanup and season analysis

    At this point in the season, we have time to clean up our code before development for code. This is important to do now so that the code remains understandable as we make many changes for worlds.

    There aren't any new features that were added during these commits. In total, there were 12 files changed, 149 additions, and 253 deletions.

    Here is a brief graph of our commit history over the past season. As you can see, there was a spike during this code refactor.

    Here is a graph of additions and deletions over the course of the season. There was also another spike during this time as we made changes.

    Next Steps

    Hopefully this cleanup will help us on our journey to worlds.

    Issues with Driving

    Issues with Driving By Cooper, Jose, BenB, Bhanaviya, Karina, and Justin

    Task: Widen Superman's wheels and plan the new robot

    Since we just qualified, we have a lot to do. On the list for tonight, between the 6 of us, we have:

  • Teaching Cooper how to write a blog post
  • Work on the model of the new robot
  • Widen the superman wheel
  • Start the bill of materials'
  • Ben and Karina worked on widening the Superman wheels by adding 2 omniwheels on either side of a newly cut shaft. This will help stabilize the robot when moving into the extended position, along with preventing falls in the future. We hope this will make it easier to drive the robot and make it more reliable. As well, we began to make the Bill of Materials for the new robot.

    Bhanaviya trained Cooper how to write and upload a blog post. Justin worked on the model for the new robot.

    Next Steps

    Next, we will work building the worlds robot.

    Planning Sessions

    Planning Sessions By Charlotte

    Task: Outline new planning sessions

    Beyond the Gantt chart and meeting logs mentioned in (T-17, Project Management), another one of the biggest additions to the team with the project management role are planning sessions. Planning sessions are a seemingly simple concept, but the team has struggled with actually implementing them. The main purpose of these sessions are to set off each member with a game plan, one that will keep them productive, engaged, and helpful to the progression of the team.

    These planning sessions take place around the main monitor in our robot house at the beginning of each practice, with a document pulled up to record our agenda. Often the whole team cannot be present, but if not the project manager reaches out to those members individually and let's them know the discussion that was had. Each session is recorded in an agenda that separates objectives into its subteam: engineering, code, blog, and miscellaneous. Each agenda is then included at the beginning of each meeting log and frequently referenced to throughout the log.

    These sessions seem like an obvious addition to the team, but we have struggled to implement this change in past years. With a project manager, there is a leading voice in these meetings that emphasizes their importance. In the future, hopefully attendance to these meetings will improve, and the whole team will recognize them as incredibly important to our success. Ways to ensure this improvement are for the project manager to create outlines before each meeting and to begin these discussions over Discord during the week in the #planning channel so that we can solidify these plans during the planning session.

    VEX 393 Motor Testing

    VEX 393 Motor Testing By Jose, Cooper, Aaron, and Janavi

    Task: Test VEX Motor 393 as a faster servo for intake

    We need to speed up our intake to spend less time in the crater collecting minerals. We can accomplish this using VEX 393 Motors with high speed gears integrated, these motors are great since they count as servos, not motors. In terms of progress, this is what we did:

    • Tested VEX Motor 393 with servo cable on BigWheel
    • Resoldered XT-30 for servo power injector cable
    • Built new cable for servo power injector
    • Did research on VEX Motor 393 Controller to find out how it works
    • Replaced gears of VEX Motor 393 with high speed gears
    • Researched how to troubleshoot VEX Motor Controllers
    We are having issues implementing these motors onto BigWheel and our troubleshooting efforts did not suffice our needs.

    Next Steps

    We need to plan how to replace the servos on the intake with the VEX 393 Motors and test their functionality.

    Balancing Robot Updates

    Balancing Robot Updates By Abhi and Ben

    Updates on Balancing Robot

    Today we managed to get our robot to balance for 30 seconds after spending about an hour tuning the PID gains. We made significant progress, but there is a flaw in our algorithm that needs to be addressed. At the moment, we have a fixed pitch that we want the robot to balance at but due to the weight distribution of the robot, forcing it to balance at some fixed setpoint will not work well and will cause it to continually oscillate around that pitch instead of maintaining it.

    To address this issue, there are a number of solutions. As mentioned in the past post, one approach is to use state space control. Though it may present a more accurate approach, it is computationally intensive and is more difficult to implement. Another solution is to set the elbow to run to a vertical angle rather than having that value preset. For this, we would need another IMU sensor on the arm and this also adds another variable to consider in our algorithm.

    To learn more about this problem, we looked into this paper developed by Harvard and MIT that used Lagrangian mechanics relate the variables combined with state space control. Lagrangian mechanics allows you to represent the physics of the robot in terms of energy rather than Newtonian forces. The main equation, the Lagrangian, is given as follows:

    To actually represent the lagrangian in terms of our problem, there is a set of differential equations which can be fed into the state space control equation. For the sake of this post, I will not list it here but refer to the paper given for more info.

    Next Steps:

    This problem will be on hold until we finish the necessary code for our robot but we have a lot of new information we can use to solve the problem.

    New Robot Base - Icarus

    New Robot Base - Icarus By Evan, Justin, Aaron, and Ethan

    Task: Build the base for the new robot

    Since BigWheel was never intended to be a competition robot, we decided to build an entire new robot based off of it. This means that the base plate of the robot is going to have to be the most accurate part of the robot since everything after that has to be built upon it. To do this, we started out by measuring the base of our original robot, then squaring the whole thing out, making sure it was uniform across the base, down to 1/32". The inner slot that houses the superman lever was done down to 1/16" because it’s precision was not as important; it houses the Superman arm's wheels.

    We cut and trimmed the basic platform using the table saw and clipped some of the thinner excess polycarb off with flush cutters. Once the base was cut to size, we moved onto the bends. The bends were measured exactly where they are on the outside of the current robot. To make precise cuts, we took a trip to the Dallas Makerspace. There, we used the sheet bender to bend our 1/8" polycarbonate which makes up the base, into shape. The walls of the base are then going to be connected to square aluminum piping that has been ripped in half to create the outer wall.

    The task of holding the sides together will be done by two 3D printed parts that will house the LED strip that goes around the outside of the robot (used to communicate to the driver which mode we are in). This base will be much more precise than our previous chassis, making it more reliable as well. Finally, the new base will have more mounting points than before, allowing for greater modularity. The old robot will be a sparring partner for driver practice. The level of craftsmanship that has gone into this baseplate is industrial grade, we have done something comparable in precision and accuracy to any product meant to be mass produced. We can only hope that our final robot works as well as it's intended.

    Next Steps

    To have a fully supported base, we need to add the framing brackets and the wheels before it can be considered a wrap on the base section of the robot.

    Finishing Icarus' Base

    Finishing Icarus' Base By Evan, Aaron, and Ethan

    Task: Perform the final steps to complete Icarus' base

    Since we finished the polycarb base, our robot went through some major changes. We last left our robot in the post-bend stage, just a piece of polycarbonate. The first thing we did was to square the whole robot with side brackets. These cleanly ripped aluminum C channel side brackets now serve as the highly accurate frame of our robot, which has been measured down the millimeter for the highest level of precision yet.

    After creating the side brackets, it was time to give them the right holes in all the right places. The holes for the rod we use as our drive shaft were drilled in the side brackets, exactly the same on either side, as were the holes for the points of attachment on either side of the robot, connecting the base to the brackets. The front bracket was cut to size and placed on the robot after the REV rail we use as an attachment point for the elbow joint was placed. Then we put the 3D printed brackets onto the REV rails that make up the back end of the frame of the robot, running the bar that became the axle for the wheels. If you want to see just how far we’ve come, you can look back at the article that Arjun and Karina wrote about building the first version of the robot over the summer. The amount of improvement is large and part of the journey. Everything on the robot is done for a reason, be it stability, weight, or efficiency. This time around we’ve significantly reduced the number of extra things on the robot, and simplified it as much as we possibly can.

    Next Steps

    The next step is going to be told in an upcoming article that will describe the process of building the arm mount. If this robot is going to be on the field and compete, it needs the elbow joint to be constructed, so that’s next on the evolution of the new robot.

    Bill of Materials

    Bill of Materials By Bhanaviya and Karina

    Task: Create a list of parts needed for the new robot

    To determine all the materials we need for the new robot, Karina and I started a Bill of Materials. To do this, we first analyzed Big Wheel sub-system by sub-system. We determined the parts used for each sub-system and placed it into a spreadsheet. Upon doing this, we needed to get each part's exact measurements so that we could save time when trying to cut the new parts. Additionally, we needed the quantity of each part as well as which manufacturer it was from. This was critical because at the end of the day, the task was to build a better version of Big Wheel but using, more or less, the same parts.

    Intake Speed

    Intake Speed By Karina

    Task: Analyze efficiency of our intake system

    A big part of our redesign is improving our intake system. To see where some of the errors may lie, we took detailed videos of our robot intaking silver and gold minerals from a side view, one mineral at a time. We measured the time between when the intake first made contact with the mineral, and when the mineral was directly underneath the rotating icecube tray, and therefore in our control, using LoggerPro video analysis.

    Silver Minerals
    TrialΔt (s)
    10.733
    20.466
    31.233
    41.934
    50.766
    60.634
    70.600
    80.466
    92.133
    100.700
    Gold Minerals
    TrialΔt (s)
    10.234
    20.532
    30.300
    40.533
    50.533
    60.300
    71.433
    80.567
    90.800
    100.433

    On average, silver mineral intake took 0.967s and gold mineral intake took 0.567s, meaning our intake was more efficient at gold mineral intake. Looking at Big Wheel intake frame by frame revealed faults in our intake. Intaking gold minerals went smoothly. For silver minerals, however, the slack in the ice cube tray resulted in it losing its grip on the mineral multiple times before the mineral was firmly grasped. This is likely the result of frictional forces struggling to overcome the elastic force of the flexible icecube tray pushing outwards. In trial 4, for example, our intake lost its grip on the mineral 4 times before it could be considered in our control.

    Next Steps: Redesign Intake Mechanism

    We are assembling a subteam of builders to take on the challenge of designing a new intake system. Some issues we'll have to address include:

    • The slack in the center of the corn-on-the-cob intake

    • The silver minerals slipping on the sorter
    • We'll have to have what changes will be made to our current design. (E-147, Intake Update)

    New Elbow

    New Elbow By Justin

    Task: Design an elbow for bigwheel that we can 3d print

    To speed up the build process of the new robot, we made a 3D printable part of the elbow joint. The design simplifies the complex assembly of the elbow mounting point and makes it a single printable part. The old elbow contains many different parts that would need to be spaced precisely in order for the gears to mesh properly, while the new print allows us to stay consistent with our measurements when building the new robot. The part contains motor mounting holes, as well as a socket to support the weight of the motor. There is also a place to put the bearing that the lift system rotates on.

    This had to be spaced properly so we calculated the exact distance by using the number of teeth and module of the gear to find the diameter. The part also has two places to attach it to a REV rail, which allows us to secure the elbow to the chassis. The spacing between the bottom REV rail socket and the bearing hole is spaced so that the gear that aligns with the bearing is flush with the front plane of the robot to stay within 18 inches. The new bearing hole is also higher up than the hole on the old robot, which gives us more extension when intaking or depositing minerals.

    Next Steps

    We need to attach the new mounts and test how the new height of the elbow mounting point affects our balance and latching.

    Updated Meetinglog Template

    Updated Meetinglog Template By Charlotte

    Task: Update the meetinglog template to more accurately reflect efforts

    An essential part of the project management role is the meeting log, where the project manager records all progress made in each subteam during each session. It requires diverse knowledge of every part of the team, and is a very important part of our engineering journal, tracking the lower level progression of the team.

    The meeting logs were previously constructed in long form paragraphs, detailing a narrative of that day in each part of the team. However, as a judge scans across the notebook, it is difficult to pick out key accomplishments from these walls of text. So, we changed the formatting of the meeting log to describe each task in a single bullet point then offering a brief feature-benefit description below said bulletpoint.

    The meeting logs are now organized as follows: Agenda (created and screenshotted then put into the log), Objective Summary (a summary of the agenda giving an overall theme for the day), Meeting Log (the actual bullet points and descriptions of the tasks that day), and the Member Meet Log (a chart at the bottom of the log that details each member present and what they worked on that day). The purpose of this organization is to allow the judge to scan the whole log and understand what we did that day, so their eyes go from broad overarching planning to specific detailed description of what we did. If they read the objective summary or look at the agenda and are interested, they are immediately drawn to the bullet points and can look at the chart in the end if they are so inclined.

    The meeting log is an ever changing addition to our notebook, and we are constantly looking to improve how our story as a team is being conveyed to judges and to people reading the blog online.

    Constructing Icarus' Elbow

    Constructing Icarus' Elbow By Evan, Aaron, and Ethan

    Task: Build the elbow for intake

    In the last Icarus' blog post, it was just getting the basic flat, support frame of the robot. The next step in the construction of Icarus' is the elbow joint that holds the intake. This time around, we simplified everything significantly as compared to BigWheel, reducing the excessive aluminum parts to two 3D printed parts. We attached these to the REV rail that runs across the front of the robot with two smaller REV rail parts we custom cut to fit the size of the 3D part. Then, we inserted the motors that each of them requires. Here we are using the same REV HD motors we used for our elbow on the last robot since they worked quite well. After inserting these, we went about supporting the elbow frame, which was done with two REV rails attached to the robot from the top of the 3D printed piece.

    These were attached at a 30-degree angle and anchored to the robot behind the two drive motors we use for the wheels. Once both of these were secured, we began assembling the arm. The arm itself has remained mostly the same, consisting of two linear slides on either side for a grand total of four, extra smooth slides. We drilled out the correct holes on all of the arm pieces, created four custom metal parts for the slides, which took a while on the bandsaw, and then assembled the bottom slide of the arm. Three holes were drilled out in four REV 86 toothed gears, which work as the mounting point of the linear slides. Once these were attached, we attached all the other necessary parts for the arm and life on the elbow joint’s 6mm hex axle that protrudes from a ½ inch hex axle set on two bearing with ½ inch hex inlay for an insanely smooth rotation. After all the necessary hardware was set in place, we put a redesigned version of our 3D printed gear keepers on to keep the distance between the motor shaft and the rotating shaft the same, and the gears firmly interlocked. During the time frame of this article, the new superman lifting lever was put into place.

    Next Steps

    The next step in the saga of the robot is the hook and the new intake, which will be seen in upcoming articles. As well, if the robot is to score at worlds, we need to construct the arm lift for the intake and then the intake itself, which will be redesigned and improved. Also, some wiring would be nice.

    Icarus' Superman Arm

    Icarus' Superman Arm By Evan, Aaron, and Ethan

    Task: Design and install a lifting arm for Icarus

    At the same time as the elbow joint was being done (which can be found in the article "Constructing Icarus' Elbow”) the Superman lift was being installed in the back half of the robot. The old superman system was difficult to install, but we designed it to be slightly easier. Mounting brackets were already pre-set in the robot so we didn’t have to disassemble half of the robot to be able to set screws into the extrusion rail. Bearings were inserted into the brackets, and the process of sliding all of the needed parts onto the rails began. First was the outside shaft collar, which holds the 6mm hex shafts in place. Then was the first interior shaft collar, which kept the internals in place. Then the first of the gearkeepers was put on, followed by a spacer meant to separate the gearkeeper’s bearing from being popped out by the gears on the Superman arm. Then came the actual Superman arm, which is one centimeter longer than our original arm, hopefully allowing more lift.

    It’s made of three 125 toothed gears from REV, with the center one’s ridges drilled out, a REV rail sized chunk sawed to insert our actual lever bar, and 3D printed spacers separating each of the gears around the outside which have all been bolted together. On the end of the bar is a 3D printed holder for the four omni-wheels we’ve positioned there, which are all set with bearings for smooth motion. Once this was slotted onto the 6mm hex rail we added one more spacer, the other gearkeeper, then the final interior shaft collar. It was put through the other bearing and bracket on the other side and finally closed off with a lost final shaft collar on the outside.

    After we got the arm in, we moved on to the driving 6mm hex shaft. Since this one was a lot longer and was hard to fit into the space provided, it was aligned in a way that it could slip through the slots of the wheels as we pushed it into place. We first put a REV core hex motor and a shaft collar that would work as the outside clamp. Then we put it into the bearing on the bracket and pushed it through. A shaft collar was placed, and then we attached the other end of the gearkeepers on. It was tight like we wanted it to be, but it didn’t make our builder lives easy. We put on a spacer to keep it in line with the Superman arm and then we put on the drive gears, three 15 tooth gears with the center one's sides cut off to mimic the Superman gears on the other side. After we put that in, we put another spacer and then the other side’s gearkeeper. This is where the struggle came. Since the gearkeepers keep the gears together exactly the distance from the center of the radius of the 15 toothed gear to the center of the 125 toothed gear, it was a very tricky squeeze to get it attached. After we managed to get it one, we put another shaft collar on and put it through the bearing on the other side. We slid on one last shaft collar on the outside, and ended the shaft with another REV core hex motor. That capped the entire subsystem off, and all that’s left is it to be wired.

    This system differentiates us from other teams - our robot is able to deposit through a lever arm that rotates the robot itself, adding an additional degree of sophistication and mobility to the robot.

    Next Steps

    The subsystem needs to be completely wired and tested before it's approved for the final robot.

    Center of Gravity calculations

    Center of Gravity calculations By Arjun

    Task: Determine equations to find robot Center of Gravity

    Because our robot tends to tip over often, we decided to start working on a dynamic anti-tip algorithm. In order to do so, we needed to be able to find the center of gravity of the robot. We did this by modeling the robot as 5 separate components, finding the center of gravity of each, and then using that to find the overall center of gravity. This will allow us to better understand when our robot is tipping programmatically.

    The five components we modeled the robot as are the main chassis, the arm, the intake, superman, and the wheels. We then assumed that each of these components had an even weight distribution, and found their individual centers of gravity. Finally, we took the weighted average of the individual centers of gravity in the ratio of the weights of each of the components.

    By having equations to find the center of gravity of our robot, we can continuously find it programmatically. Because of this, we can take corrective action to prevent tipping earlier than we would be able to by just looking at the IMU angle of our robot.

    Next Steps

    We now need to implement these equations in the code for our robot, so we can actually use them.

    Icarus' Arms

    Icarus' Arms By Evan, Aaron, and Ethan

    Task: Install intake arms

    Since the last post, in which we installed the Superman Arm, we've installed the second stage of the linear lift and the belt drive that accompanies it. We began by drilling two holes in the linear slides that were exactly the space between the holes on the carriages for the linear slides using a drilling template we printed on the Tazbot printer. We did this to two of our linear slides, and then attached them. We realized that they were too long and sticking out of the 18x18x18 sizing box, so we detached them and cut off a centimeter from the top and ground off the edges. They were reattached successfully, and the 3D mounts for the belt system were installed at the same time since they use the same point of attachment as the linear slides.

    Those custom pieces that were mentioned in the Joint Operation article were now utilized, attaching to the top of the first linear slide and to the carriage of the second linear slide. These parts are used for the attachment of the pulley bearings that the belt drive relies on to function. We installed these pieces rather easily but struggled on some of the tighter fits that were done to reduce wiggling in the arms, a problem that the last robot had. The next thing we added was the physical belt which drives our lift. The belt was tied off on the final carriage on the second linear slide on either side. The next step was to create the mounting for the motors that would drive the lift. To do this we set up a REV rail under each of the elbow motors, and then topped it off with another rev rail that we connected to the elbow frame supports that run from the front to the back of the robot. Then we mounted the motors, two Orbital 20 andymark motors, which at first didn't fit. The issue was that there was no way to mount them close enough for a belt to be put in place with the current gear keepers we had on the robot. They were attached, and then the motors were mounted, and the belts were put on. The lift has the same ratio as last time, which is further explored in the article Bigwheel Upgrades. The whole system is much more cleaned up and simplified, and generally looks a lot better.

    Next Steps

    The next challenge for us is going to be making the hook, attaching said hook, and redesigning the intake in time for effective driver practice.

    Project Management Mentorship

    Project Management Mentorship By Charlotte

    Task: Ensure skills are passed to underclassmen

    Since our project manager is leaving for college next year, there has been an effort to teach the younger students on our team to take on this role and its many responsibilities. These responsibilities include updating the Gantt chart, writing meeting logs, gathering information for meeting logs when you are not able to make it to meetings, leading and helping writing post mortem and roads to, ensuring general organization for the whole team in terms of Discord and other communication methods, writing articles about the ever-changing responsibilities of this role, managing competition day roles and management, leading and recording planning sessions, being part of leadership in the blog sub team, ensuring communication between the various subteams in Iron Reign, encouraging and understanding detailed explanations of each part of the robot, blog, code, and presentation, among much more.

    This is a lot for one person to take on, emphasizing the importance of gentle and detailed mentorship so that next year our new project manager has all the tools and much needed coaching they need to succeed and not get lost in what the role entails so that they can make the team a more organized unit.

    We have taken on many freshmen interested in assuming these responsibilities, notably Bhanaviya and Cooper. This mentorship begins with the meeting logs, which often take multiple hours to construct due to the fact that they must understand not only what each member of the team is working on, but also how that plays in the overall progression of the team. One big example is in conveying the progress of the coding team. This has been a challenge for me this year due to my lack of experience in dealing with robot code. Taking the time to have a longer discussion the the coders and demanding explicit details about the code changes and how these changes affect the overall progression of the code is what helped me with this challenge. This demand for detail is what is most important in the mentorship process, as it takes a certain confidence and assertion to do so.

    Aside from these soft skills, there are some hard skills to be had too. First of all, we mentored all the underclassmen on how to use HTML to write and post a blog post as well as an introduction to what their language should sound like in these blog posts. Rather than conversational, each post should be written in a professional, technical, or formal manner, depending on the subject matter of the post. Meeting logs have their own template and formatting, which have been taught to future project managers so that they can practice these skills. Bhanaviya has already written a promising number of meeting logs with impressive detail.

    As the season comes to an end, there a few things remaining to teach, especially planning sessions and the Gantt chart. The Gantt chart especially requires a lot of hands-on mentorship, as though the software is intuitive it is difficult to be in the mindset for that type of higher level organization if you haven't ever before and haven't been walked through it. Alongside this mentorship, I will have the freshmen lead planning sessions with me as an advocate alongside them, so if the conversation gets off topic I can supply them the confidence needed to call the meeting back to focus. Mentorship is a long process, but is essential in such an abstract role in the team and I will continue to be there as a voice of support throughout the whole process.

    Intake Flappers

    Intake Flappers By Jose, Evan, and Abhi

    Task: Design and test intake flappers to speed up mineral intake

    Due to our new intake articulation involving the superman wheel the ice cube tray intake is slightly too elevated to intake minerals. To fix this we designed small flappers out of ninja flex(the Iron Reign way) to help the intake reach further. Tests prove this intake to be quicker than the ice cube tray alone and it should suffice for the UIL State championship tommorow.

    Next Steps

    We will compete at UIL and see if the new intake works

    UIL 2019

    UIL 2019 By Ethan, Charlotte, Evan, Janavi, Beno, Benb, Bhanaviya, Abhi, Arjun, Jose, Aaron, Paul, Cooper, and Justin

    Task: Compete at the Texas State Championship

    Today, we competed at the Texas State Championship, UIL Robotics, Division 5A-6A. We finished our robot earlier this week, so this served as a testing ground for our new robot and code.

    Judging and Awards

    There is no presentation at UIL - the judges appear at the pit ad-hoc to ask questions. And, there are no real awards. In this case, we talked to the judges, and they enjoyed our robot, but they happened to watch the game where our robot failed to move due to the gears breaking, so we were not under consideration for any awards.

    Talking to BAE Systems

    Usually at UIl there is a special aisle dedicated to visiting colleges and companies who support FTC teams and want to watch the competition. This time one of the visiting compaines was BAE Systems. Janavi went and talked to one of their employees who was able to connect her to the Dallas team. We plan to contact them to learn more about how they use the conecpts we are learning their jobs. We also hope to be able to give them our presentation and a run down of our robot and its capabilites.

    Code/Robot/Robot Game

    As the robot was freshly built, we didn't have much coded before the tournament. The night before, we did some basic tuning and created an autonomous, but not much. This coding is detailed in an earlier post. Despite this, the autonomous performed reasonably well - we could reliably delatch and sample - our issues came up in scoring the team marker as we failed to consider that the team marker wouldn't fit in the redesigned intake.

    The tournament also served as a stress test for Icarus. Two major issues cropped up: the belt system and the Superman arm. First, the belt system itself worked well - Icarus' arm extended quickly, but it repeatedly got caught on the lander's edge, detensioning the belt and requiring constant maintenance. Second, the gears on the Superman arm were stripped as we attempted to escape the crater in our first match. The stripping itself isn't surprising - Superman applies pressure on the gears' teeth on the order of mega-Pascals, but the quickness of stripping implies that the gears of Icarus do not fit together as well as BigWheel. So far, we plan to redesign the Superman arm with metal gears to reduce the stripping.

    Game 1
    We won. Our autonomous worked perfectly, but we overshot the crater while parking and got stuck (this was due to underestimating the speed of the 20's on our robot). Thus, we were completely stuck during teleOp, but our partner carried us.
    Game 2
    We lost. When we put the robot on the field, we realized that Superman's gears had stripped, but it was too late to change them out. So, we were stranded in the middle of autonomous and couldn't move beyond that.
    Game 3
    We lost. We hadn't fully repaired Superman, so we were again stranded on the field.
    Game 4
    We lost. We set up an untested autonomous, creating a point deficit we couldn't recover from.
    Game 5
    We won. Superman was fixed and our autonomous worked allowing us to pull ahead by 20 points and win the match.

    Next Steps

    These will be detailed in the UIL post-mortem.

    UIL 2019 Postmortem

    UIL 2019 Postmortem By Ethan, Charlotte, Evan, Janavi, Beno, Benb, Bhanaviya, Abhi, Arjun, Jose, Aaron, Paul, Cooper, and Justin

    Task: Reflect on what we did correctly and incorrectly at UIL

    Pit & Packing & Roles

    • Pack more robot parts - didn't have enough to repair Superman arm
    • Pack more tools - needed soldering iron to repair voltmeter
    • Better organizational system - we couldn't find tools easily
    • Need handouts - see tokens post
    • Need team visibility - get shirts for freshmen, get people in stands
    • Need responsibility for clean pit - messy pit made robots repairs much harder
    • Need preassigned roles for team members on game day - reduce confusion
    • Need better scouting system - use Google Forms and live scouting

    Robot & Game

    • Need to repair Superman arm - gears stripped in match; will replace with metal gears
    • Need to install linear slide belt protector - belts got stuck on lander
    • Intake needs to be clear - remove friction tape
    • Need to reduce sorter bar in intake - reduces visibility
    • Need driver practice - reduce simple errors
    • Need auto setup practice - reduce simple errors
    • Need new team marker - old one did not fit in intake

    Code

    • Need to enhance lights system for teleOp - better driver knowledge
    • Need to calibrate anti-tipping method - not adapted for Icarus
    • Need to slow crater-side auto - prevent crater parking mishaps
    • Need to calibrate depot-side auto - options when working with other teams
    • Need to find Superman-linear slide equation - easier articulations
    • Need to simplify controls - automate intake, deposit for driver accessibility

    New Superman Arm

    New Superman Arm By Ethan and Evan

    Task: Redesign the Superman arm to be more robust for Worlds

    In posts E-116, we found that we were putting pressure on the individual teeth of the Superman gears on the order of mPa. We designed gearkeepers to ensure that the gears would interlock and reduce pressure, and these worked for awhile. However, under tournament pressures at UIL, the teeth on the smaller gears broke entirely - between the teeth that composed the gearing-up portion, at the beginning we had 45. At the end, we had 15 teeth.

    This necessitated a total redesign. Upon coming back from UIL, we created a new version of Superman with metal Tetrix gears with a 3:1 ratio - the aluminum Tetrix uses has proven much tougher in the past. To compensate for the reduction in gear ratio, we removed the old Core Hex Motors and replaced them an NeverRest+BaneBots 104:1 motor+gearbox combination. Coming off the bat, the NeverRest outputs .17 N*m, and with the gearbox, it outputs .17*104=17.68 N*m. With the 3:1 gear ratio, it outputs 53 N*m, matching the previous Superman arm while increasing tooth durability.

    This new Superman arm will allow us to rotate the entire body of our robot around the axis of its wheels, allowing us to reach the lander without difficulty and ensure redundancy on the robot. The Superman arm is the centerpiece of our robot; it allows us to utilize Balancing, Center of Gravity Calculations, and Articulations in a truly innovative way.

    Next Steps

    We need to test the arm to make sure no additional stripping occurs.

    Intake Update

    Intake Update By Ethan

    Task: Custom design an intake to improve intake times

    In testing, we found that the intake didn't perform adequately - the balls would slide back out in the inverse articulations. So, we designed attachments for the corn-cob intake out of ninjaflex, figuring that small tabs would hold the minerals in better. It failed - they were too compliant - but we found it was much easier to intake minerals than before due to the high coefficient of friction.

    So, we decided that the corncob base was the issue. We designed a circle with the diameter of the previous corncob aligners and attached thicker tabs on the outside, creating the stl seen above. When tested, this was much less compliant than the previous beater bar, which served to make intake easier. In addition, the combination of reinforced tabs and ninjaflex prevented the minerals from falling out of the intake through increased coefficient of friction.

    Next Steps

    We plan to reattach this to the robot to do driver practice.

    Machining Gears for Superman

    Machining Gears for Superman By Ethan and Justin

    Task: Machine replacement gears for Superman

    Shortly after creating the new Tetrix gear system, we got a response from one of the CNC shops we'd reached out to, offering to machine the 15 and 125-tooth REV gears from the STEP files. So, we took the Superman system off of our old robot, BigWheel, and sent some of the broken 15-tooth gears from UIL.

    In response, the shop sent us the new gears the next day, with added modifications for mounting the gears onto REV extrusion. These gears will make the arm much stronger, making it more robust and able to withstand the shear pressure on the teeth.

    Next Steps

    We need to mount the gears and test them to ensure stability.

    Ninja Flex Intake V2

    Ninja Flex Intake V2 By Jose, BenB, Karina, Evan, Abhi, Ethan, Charlotte, and Aaron

    Task: Design, implement, and test a newer version of the ninja flex intake

    The new ninja flex intake is good, but it has room for improvement. One issue is that it is too big and minerals have some problems entering the intake tray, Another issue is that the spacing of intake gears is too much and cuases minerals to be intaked slower. We fixed this by using smaller intake gears and using six of them instead of five. After replacing them we could test the new and improved intake. Results showed a much faster intake speed with an average intake time of 1-2 seconds. This was a major improvement and most likely the intake's final iteration.

    Next Steps

    Now with a finished intake we can drive test to see its functionality in a real match.

    Final Gantt Chart

    Final Gantt Chart By think

    Task: Update the Gantt Chart

    Earlier in the year, we posted an early version of the Gantt chart as seen in (T-17, Project Management). Since then, the chart has seen many changes, which can be seen below:

    See finished Gantt chart at front of notebook in pocket.

    Since the last update, we have added a few groups, notably research and development. The Gantt chart, along with other higher-level planning is completely foreign to the team, so it has been a journey to accomplish this progress. This year was a test of the concept, so next year we will work to improve on this concept and expand its use from strictly the project manager to the whole team. Expect to see another Gantt chart next year that is more fleshed out, detailed, and accurate.

    Control Hub First Impressions

    Control Hub First Impressions By Arjun and Abhi

    Task: Test the REV Control Hub ahead of the REV trial

    Iron Reign was recently selected to attend a REV Control Hub trial along with select other teams in the region. We wanted to do this so that we could get a good look at the control system that FTC would likely be switching to in the near future, as well as get another chance to test our robot in tournament conditions before Worlds.

    We received our Control Hub a few days ago, and today we started testing it. We noticed that while the control hub seemed to use the same exterior as the First Global control hubs, it seems to be different on the inside. For example, in the port labeled Micro USB, there was a USB C connector. We are glad that REV listened to us teams and made this change, as switching to USB C means that there will be less wear and tear on the port. The other ports included are a Mini USB port (we don't know what it is for), an HDMI port should we ever need to view the screen of the Control Hub, and two USB ports, presumably for Webcams and other accessories. The inclusion of 2 USB ports means that a USB Hub is no longer needed. One port appears to be USB 2.0, while the other appears to be USB 3.0.

    Getting started with programming it was quite easy. We tested using Android Studio, but both OnBot Java and Blocks should be able to work fine as we were able to access the programming webpage. We just plugged the battery in to the Control Hub, and then connected it to a computer via the provided USB C cable. The Control Hub immediately showed up in ADB. (Of course, if you forget to plug in the battery like we did at first, you won't be able to program it.)

    REV provided us with a separate SDK to use to program the Control Hub. Unfortunately, we are not allowed to redistribute it. We did note however, that much of the visible internals look the same. We performed a diff between the original ftc_app's FtcRobotControllerActivity.java and the one in the new Control Hub SDK, and saw nothing notable except for mentions of permissions such as Read/Write External Storage Devices, and Access Camera. These permissions look reminiscent of standard Android permissions, and is likely accounting for the fact that you can't accept permissions on a device without a screen.

    While testing it, we didn't have time to copy over our entire codebase, so we made a quick OpMode that moved one wheel of one of our old robots. Because the provided SDK is almost identical to ftc_app, no changes were needed to the existing sample OpModes. We successfully tested our OpMode, proving that it works fine with the new system.

    Pairing the DS phone to the Control Hub was very quick with no hurdles, just requiring us to select "Control Hub" as the pairing method, and connect to the hub's Wifi network. We were told that for the purposes of this test, the WiFi password was "password". This worked, but we hope that REV changes this in the future, as this means that other malicious teams can connect to our Control Hub too.

    We also tested ADB Wireless Debugging. We connected to the Control Hub Wifi through our laptop, and then made it listen for ADB connections over the network via adb tcpip 5555. However, since the Control Hub doesn't use Wifi Direct, we were unable to connect to it via adb connect 192.168.49.1:5555. The reason for this is that the ip address 192.168.49.1 is used mainly by devices for Wifi Direct. We saw that our Control Hub used 192.168.43.1 instead (using the ip route command on Linux, or ipconfig if you are on Windows). We aren't sure if the address 192.168.43.1 is the same for all Control Hubs, or if it is different per control hub. After finding this ip address, we connected via adb connect 192.168.43.1:5555. ADB worked as expected following that command.

    Next Steps

    Overall, our testing was a success. We hope to perform further testing before we attend the REV test on Saturday. We would like to test using Webcams, OpenCV, libraries such as FtcDashboard, and more.

    We will be posting a form where you can let us know about things you would like us to test. Stay tuned for that!

    Project Management Post-Mortem

    Project Management Post-Mortem By Charlotte

    Task: Evaluate the Project Manager position

    This year, I started the role of project manager, and there have certainly been plenty of growing pains. Iron Reign had previously learned to embrace chaos, frequently pulling all nighters and fumbling to finish each part of the robot in a timely manner. In this post, I will discuss all of the different aspects to being a project manager on Iron Reign so that we can continue to improve on our organization. The main focus will be the meeting logs, planning sessions, and the Gantt chart.

    • Meeting Logs
    • This year we have completely changed meeting logs. We changed the format to using bullet points rather than long-form, and the way they are told to using feature-benefit language. Feature-benefit describes the what (taking up 2/3rds of the description) followed by the why (taking up 1/3rd of the description). These descriptions are incredibly important to concisely portray our progression to the judges.
    • Planning sessions
    • In previous years, we have had trouble implementing these planning sessions effectively and we still have this difficulty. When there is no project manager present, they don't occur at all and must be supplemented with discussion in the Discord. They have been very helpful in constructing agendas in meeting logs, but next year we are going to push the need for these sessions even more. They make sure that every member has a task to accomplish during the meeting and help remove the pull of distractions. In mentoring the freshmen, hopefully these needs will be met.
    • Gantt chart
    • The Gantt chart has been the most difficult factor of project management to implement. The higher-level organizational mindset required is one difficult to acquire without any close mentorship. Our Gantt chart has seen many changes, especially earlier in the seasom, but dropped off nearing the end of the season due to other responsibilities. Next year, the most important improvement would be to involve other team members in the creation of the chart a lot more than this year. This would help the chart accomplish higher detail and accuracy as well as allow it to be helpful and references by not only the project manager, but the whole team. They will be implemented into the planning sessions

    Next year, there are quite a few improvements to be seen in this role. This was the first year and going in with no previous experience and with the team not used to such a role has been a challenge. Hopefully, most of the mistakes to be made have already been made, and the project manager role in the team can be seen as important to the organization and overall well-being of the team. It requires intense dedication, confidence, and organization, which I have tried my hardest to provide this year, but I have faith that with the amazing abilities of our team, we will improve our organization and project management for years to come.