FTC 2019 - 2020
2020's fourth robot, Storm, was built to compete in FIRST's challenge: SKYSTONE presented by Qualcomm. Our robot combines input from 6 distance sensors, 5 encoders, and a camera—along with a custom computer vision algorithm—to achieve an autonomous program that can score 2 stones(4''/5''/8" plastic lego-shaped blocks) in autonomous, move the foundation, and park. In Teleop, our robot was able to stack up to 5 stones high.
This year for our drivetrain we wanted to move away from our standard tile runner chassis we have used for the past four years, and augment our own. A typical tile runner chassis has motors which stick far inside the drive train, but we wanted a pass through robot this year. For that reason, we custom laser cut our plates with the help of FTC Team 14436 Roboctopi, and then mounted the motors according to our needs, giving the stone space to pass through. We continue to use Mecanum wheels with a 1:1 sprocket ratio and 20:1 motors to allow for strafing, speed, but also strength as compared to omni-wheels.
The depositor has gone through many iterations, so I will highlight how we chose our design, and used the engineering design process to help. We first used LEGOs to prototype our nub gripping design, that would apply pressure to the back of the stone and grip the back nub. We discovered this would take too much precision, so we needed a more general claw that could grab the sides. Our first one of these claws was found in Fall, and used one servo to power the 2 side plates. We used CAD to plan out how this would assemble, and we completely 3d printed it so that we would have the maximum effectiveness, by making specialized pieces.
Our collector has gone through many iterations. The original, simple collector on Puddle used two 4-inch compliant wheels to cycle stones onto the foundation. The second collector, found on River, was placed on an extension so that we could reach across the field to the depot. Though it was a good idea in theory, the collector was not consistently effective due to its size and stringing problems. Our final collector is found on Fall. Fall’s collector uses two “‘flappy arms” that have 2 inch stealth wheels as well as 2 inch compliant wheels to most effectively grip the stone and bring it into the range of the depositor. This has been the best collector because of our effective engineering research and iterations.
Three main iterations were made on our lift. Our first iteration took place on our League Meet 1 robot, River. On River we actually had two lifts, one horizontal and the other vertical. While both systems worked perpendicularly, they both followed the same design. They were REV bars connected to each other via the REV linear extension sliders and custom 3-D printed pulleys on the ends. The stringing was done in a cascading design where there was one string in between two slides and one both sides. We also used REV Hex Motors to power the slides. Through testing and practice we learned that REV Hex Motors are not the best for high torque needs especially when there is only one to do the task. This led us to our next lift which was on Fall. The lift on fall used continuous stringing with 2 neverest 20:1 orbitals. This proved to work well, because of our string tensioners- an innovation that kept our string constantly tight, for the continuous rigging. On our third iteration, we removed a lift motor because after further calculation we realized we didn’t need it. With an extra motor in hand, we were able to replace the horizontal extension servo so it can run faster.