Articles by tag: think

Articles by tag: think

    Swerve Drive Prototype

    Swerve Drive Prototype By Abhi and Christian

    Task: Build a Swerve Drive base

    During the discussion about swerve drive, Imperial robotics, our sister team, was also interested in the designs. Since we needed to conserve resources and prototype, 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. This system would be mounted to a square base. We decided to go ahead and make the base.

    Immediatly 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 train 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.

    Next Steps

    Wait until Rover Ruckus starts so that we can think of a new chassis.

    CNC Machine Rehab 1

    CNC Machine Rehab 1 By Ethan and Charlotte

    Task: Refurbish an Apple II CNC Mill and Lathe Set

    We were helping our school's FRC team clean out their parts closet, which hadn't been cleaned in 10-ish years. Under the layers and layers of FRC junk, we found an Apple II-operated Patterson/Paxton CNC Milling Set. These were meant to run off of a long-since-gone Apple II in a classroom setting. But, it had long been auctioned off, leaving the set useless. But, Iron Reign, as a collective of hoarders, decided to bring these machines over to the house to refurbish.

    The first idea we looked at was emulating the Apple II with an Arduino, as seen here. However, this implementation didn't have the response rate needed for an accurate CNC machine, so we scrapped it. Then, we found this post. The problem that people mainly encounter is that, for some strange reason, Paxton\Patterson used a proprietary parallel port pinout, and deviating from that pinout (read: using a standard parallel cord) would fry the optidriver board in the machine. So, we bought a ethernet-to-parallel port jumper box (UC300eth).

    We then sliced a parallel cable in half, and rewired the wires to the pins, treating the left column of that of the port numbers on the board and the right as the pin numbers of the cables.



    We then made a power supply for the UC300eth. We attempted to use a 10V DC power supply, and use a voltage splitter. Unfortunately, the power spiked, and probably fried the UC300.

    Next Steps

    We need to buy a new UC300 board and hook it up to a laptop with Mach3 to test the power.

    Technicbots Chassis Project - July Meeting

    Technicbots 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 Technicbots' 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.

    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 mamograms. 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.

    Relic Recovery Brainstorming & Initial Thoughts

    Relic Recovery 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 codebase 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, Evan, Aaron, and Freshmen as to be determined

    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 seperating 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
    Claiming Autonomous 15 1 High
    Parking Autonomous 10 1 High
    Sampling Autonomous 25 2 Medium
    Latching End Game 50 3 High
    Robot in Crater End Game 15/25 1 High
    Mining [Depot] Tele-Op 2 per item 1 High
    Mining [Cargo] Tele-Op 5 per item 2 High

    Brainstorming Two - Enter the Void

    Brainstorming Two - Enter the Void By Evan, Abhi, and Janavi

    Task: Have a 2nd brainstorming session

    Last week, we had a lot of new recruits show up for the FTC kickoff. In fact, a bit too many. Luckily for us, we either scared them off or they realized that they'd like to move to FRC. So, today's session was a bit more managable, and we were able to break down into some new building tasks.

    Intake System 3 - TSA Bag Scanner

    If any of y'all have ever been on an airplane, you've gone through airport security. This part of our robot is inspired by the bag-scanning machine, 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.

    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. So, as the true hipsters of FIRST Tech Challenge, we want to move onto something new and fresh.

    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 seperator, 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.