Team 7762 AutoPilots: 2021 Robot Reveal - Maverick III

Team 7762: AutoPilots is proud to present our third robot: Maverick III

Maverick III Reveal

Maverick III is a direct descendant of our 2020 robot Maverick II. After finishing as finalists in week one at FIM MCC in 2020, we knew we had a solid foundation to build upon, with newfound opportunities to train new members and undertake new challenges alike. Thus through collaboration and rigorous testing Maverick II had a complete overhaul, in order for Maverick III to take flight.

Features of Maverick III
Drivetrain: KOP Andymark 6 wheel drop center tank-drive chassis, with Falcon 500’s swapped in place of the CIMS. We knew we needed power while reducing weight, so Falcon’s were the obvious choice.
Intake: 2 Roller System Driven by a 7:1 775pro, Allows for the Robot to drive at max speed and intake at the same time
Bottom Roller: PVC based Roller wrapped in MotorCycle HandeBar Foam
Top Roller: Steel Bar Roller that is chained with PolyUrthane Belt to our Bottom Roller. This Roller’s intention is to maintain control of the Power Cells at all times, giving our intake control over them from initial touch to its transition into our indexer. Our Top Roller also features a system we call Mav Shell which is a wall of duct tape and hockey tape. We spent over 25 hours testing our intake, intaking a total of over 2500 power cells in order to ensure success. Through this we found out that the PolyUrthane Belt likes to ride up to the highest point of rotation or contact possible, no matter how tight you make the belts(We bent our intake which is 1 inch thick rectangular tubing). Thus by making eggshell shapes out of tape, as the belt begins to slip, it will naturally ride back up to the center of the eggshell, thus allowing our intake to be used at any angle without worry.

Hopper/Index
Our Hopper is a passively designed mechanism, and one of our simplest yet most effective parts of the entire robot. After having issues with static discharge last year, which caused our robot to be half wrapped in tin foil, and required the assistance of the lead robot inspector for FIM to solve it, we decided to move away from belts and motors(KISS am I right). Thus we created a ramp on a 10-degree slope. The ramp begins at 27 inches wide and leads into an 8-inch wide funnel, leaving 1/4 inch of room on both sides of the power cell.
Our indexer is a 4-inch compliant wheel, powered by a 775 pro. This is powered down in code to 50 percent output allowing 4 volts of current to go through, which is the maximum load that the given 775 pro can run at before stalling and eventually causing a fire.

Feeder:
Our feeder is a continuation of our hopper, following the same 8-inch wide rule. Our Feeder is driven by a Falcon 500, which is directly driven at 100 percent output and features two hex shafts connected by a polyurethane belt, and of course, our MavShell Technology.

Shooter:
Our shooter is a hooded flywheel shooter, direct driven by a Falcon 500, with adjustable RPMs instead of an adjustable hood. Preset at an angle of 37 degrees, with a 1.5-inch compression throughout, we are able to hit the inner port in every zone, but the green zone, consistently. For the flywheel itself, it consists of two 4 inch colsan wheels slapped right next to each other, allowing the power cell to ride to the highest point, which is directly in between the two. Our flywheel also consists of four SDS brass flywheels for weight, bringing our total flywheel weight to be 7.5 pounds with an exit velocity of around 50 mph. Furthermore, our shooter features Closed Velocity Looping, smart feeding, PID auto-alignment, and vision processing.

Software:
Good software turns a good robot into a great robot, and we pride ourselves on software every year. Therefore we wanted to make our shooter the best it could possibly be.

Smart Feed: Our Smart Feed system is nothing extremely special, but it is noteworthy as it allowed our shots to be extremely consistent. By using the Falcon’s built-in encoder, and Phoneix Tuner’s Self Test Snapshot, we were able to create a threshold-based system within code. How this works is by knowing the max rpm at our given location, we can adjust our feeder to switch on and off, by using thresholds. These thresholds tell the robot that if it is greater than a 300 unit difference the feeder falcon must switch off, and if there is less than a 50 unit difference, to turn on. This effectively creates a buffer zone of around 75 rpm. We know that every time we shoot a power cell, our flywheels rpm drops. Thus by switching our feeder on/off until our flywheel is up to speed, we can ensure the power cells leave at the same velocity every time.

RPM Control- Our shooter is programmed to the four zones of the challenge and with the press of a button, changes the rpm of the Falcon in order to hit the inner port. This also works with our PID System.

Vision - Using a Pigeon IMU and Limelight 2+ we created a PID-based system, tuned to the different zones. We found that certain zones saw shots land constantly on one side of the port the farther we got from the port. This is due to the offset of our limelight. To combat this we implemented a zone-based offset feature. When the operator changes the rpm of the shooter according to the zone the robot is in, it relays to the limelight that it needs a different PID. Therefore when the driver enters PID mode, the robot adjusts its heading to automatically align to the center of the port. This greatly aided in our ability to hit the inner port, and when completed had a maximum auto-align angle of 45 degrees, meaning that as long as part of the port was seen, the driver could enter PID mode with 45 degrees of error and the robot would align to the center of the port in under 1.5 seconds.

Drivetrain Adjustments- This is arguably my favorite part of the entire robot, and maybe yours too. Maverick III has ABS braking systems. Yes, you read that right, a robot has ABS braking. Why is this good though? We found out that Maverick III had a lot of power, so much so that we could not drive on the carpet without doing a burnout or drifting, and if it touches tile it turns into Toyoko Drift(it’s pretty cool, to be honest). Burnouts are bad as it a) takes time and b) is not efficient. Therefore, we needed a solution, and we did not want to reduce the output of the motors. Therefore we added stator limiters, effectively reducing burnouts and drifting to near zero, and instead of having to mimic “pumping the brakes” our robot did it for us, and overcame it. The greatest part is that these can be adjusted to the surface we are on, allowing us to maintain control over Maverick wherever we go, and in the future allow us to take our robot to show off at places, and not worry about being unable to go at 100 percent speed.

If you have any questions at all please feel free to leave them in the comments below!!

6 Likes

I love this robot. Super simple, very effective. Awesome job!

Thank you!! That was our whole goal, simple and effective solutions to problems allow more time for actual testing and fine-tuning, and less time spent fixing one stubborn piece of the robot.