Friday, April 14, 2023

Event Recap: Texas State Championships 2023

Results:

Award: Gracious Professionalism Award


This past weekend, Spectrum competed in the FIT District State Championship. For playoffs, our alliance partners were FRC#5427 Steel Talons and  FRC#2582 PantherBots.


Spectrum had a 9-3 win-loss ratio in Qualifications. During playoffs, we lost our first match in the upper bracket, and continued to lose in our second match against 5415 Pearadox, 5431 Titan Robotics, and 9140 The Robobots, which ultimately eliminated us from the competition. 




Failures and Fixes:

Failure: The charge station couldn’t press down in a few matches.

Fix: Raising the rear bumper a small amount seemed to help but the charge stations were still not as smooth as in previous events.


Failure: Pulley flange on the inside roller failed at a print layer and fell off

Fix: Replaced the pulley

Failure: Elevator lower idlers bend at the shaft. The autonomous mode was stalling down the elevator and that was a contributing factor to it bending

Fix: Replaced, and have no solution that mounts them together for Champs


Failure: The elevator belt had a tear, which we caught during visual inspection before it failed

Fix: Replaced the elevator belt


Failure: Damaged the tread on one of our wheels. Because the replacement tread was significantly bigger in diameter compared to the original diameter, the discrepancy between the new and old diameter would affect our autos. This is the first tread swap of the season after the damaged one from Waco.

Fix: Swap treads on all four wheels with new ones 


Failure: Misaligned bumper mounts to the robot.

Fix: Bolting the ones that were aligned and it was fine for the rest of the event


Controls Update

  1. Implemented AprilTag relative aiming for cube node scoring. The field pose estimation wasn’t accurate at our first two events. We made quick control loops to center us on the cube nodes for cube scoring.

  2. The cone scoring command retracts the 4 bar and lowers the elevator at the same time, making that process quick and smooth compared to the districts.

  3. Had high cone auton modes on clean and bump sides.


Planned Improvements:

  1. Improved autonomous modes

  2. Change to cube scoring to become more consistent.

  3. Flashlight to help with mid-field intaking


Strategy and Scouting Recap:

  • Our scouting went well. Our data was pretty accurate, and most of our match and alliance predictions were true. 


Thank Yous:

  • Huge thank you to our alliance partners: FRC#5427 Steel Talons and  FRC#2582 PantherBots; you guys were awesome to cooperate with! We wish you a happy rest of the season.


  • Also huge thank you to our sponsors: Analog Devices, Solarcraft, QAD Works, Team REV, Gene HAAS Foundation, ISA - Houston Section, Texas Workforce Commission, and Intuitive Foundation! You all made this brilliant season possible.














 

Thursday, March 30, 2023

Event Recap: Houston District 2023

Results:

Awards: Winner, Gracious Professionalism Award, and Dean's list semi-finalist (Photon 8515).


This weekend Spectrum competed at the Houston District Event. We would first like to thank our alliance partners: 118 Robonauts, 2585 Impact, and 9121 Wild West.


Spectrum went 15-4 overall at the event, finished ranked 3rd, and was invited by the 1 seed, 118, to join their alliance. After inviting 2585 (and later 9121) to join this alliance, we lost our second match in the upper bracket and made it to the finals through the lower bracket, winning 2-0.

Our development subteam, Photon 8515, also competed. Photon went 5-9-1 overall at the event, finished ranked 21st, and was the first pick of the 5th alliance captain, 2587 alongside 7691. They lost match 2, won match 5 and were eliminated in match 10.





Failures and Fixes:

Failure: flange on the metal launcher pulley came off.

Fix: Added a 22-tooth ~1/2 in 3D printed pulley as a temporary spacer to keep the flange from coming completely off.


Failure: sponsor panel zip ties breaking and flapping around as we move

Fix: bent sponsor panel behind 1x1 REV max tube; future fix is to add rivet nuts for sponsor panel.


Failure: rollers not going at correct velocity (motors would stall and not intake cubes well).

Fix: change velocity control to percentage control.


Failure: we experienced out of memory errors during practice matches. 

Fix: loaded fewer path planner paths.


Failure: bumper mounts broke.

Fix: replace bumper mounts and rivet to the back (note: this was a bad idea, don't do it again, takes too long to drill out); to be done: metal brackets?


Failure: limelight field pose estimates off by 10-20cm.

Fix: none, turned it off completely. :(


Failure: cable carrier broke.

Fix: zip-tie cables to carriage.


Failure: swerve steering belt tooth sheared off.

Fix: temporary: spam reset swerve to absolute the button during the match to keep the wheel pointing the right way even when the belt skipped. Permanent: replace belt; check how much belts have worn on the swerve

Failure: improper strain relief. Wagos got tugged and came undone because they were under stress.

Fix: to be done: redo wago configurations with better strain relief, more zip ties.


Failure: lower intake rollers not spinning; belt partially dislodged.

Fix: we discovered after the match that pressing eject would fix the problem. A more permanent solution is replacing the 3D printed pulley with a version with a larger inside flange so that the belt cannot possibly slip over it.


Controls Update:


Autonomous:

We ran a three-ball and angle/balance for the first day of the event. Autobalance was too slow sometimes, and the auton period would end before it got itself balanced. We fixed autobalance in time for matches on day 2 and ran a three ball and balance sequence for the rest of the event.


Planned Improvements:


  • Improve cable carrier

  • Raise launcher wheel to reduce angle and reduce accidental cube launches.

  • Improve bumper brackets

  • Software improvements

    • Make our stick-steer command only turn to the cardinal directions

    • Make our cone scoring button start a full routine that brings the elevator down

    • Add a pilot button to shift the robot left or right roughly ½ a cone distance for easier alignment to cone nodes.

    • Ensure that slow mode and FPV mode work with heading lock

    • Get basic limelight alignment to cube nodes working (no longer trying to use field pose estimation)

    • Work on making sure our 3 cube autons actually score all 3 cubes.

Strategy and Scouting Recap:


  • Scouting data was better at this event than Waco. The strategy and scouting team did an excellent job gathering robot photos and match data for each robot. We also were able to have good match preview data for all our matches to better plan strategy with our partners.


  • Our scouting meeting was relatively short: about 2 hours to get us a full 24 team list with tiers to prepare for alliance selection.


Thank Yous:

  • Thank you to Robonauts 118 for selecting us to be a part of the #1 alliance. It was great working with you again; good luck this weekend at Space City District.

  • Thank you to our other alliance partners 2585 Impact and 9121 Wild West; you all were excellent to work with and we hope to work with you in the future.


Houston Photos