Team 5012 Offseason Reveal

[Jeremy Germita]

Originally Posted by kylemikableh

Looks great! How is that pixy? Is it worth getting for future image processing?

We love the pixy. Rev1 of our rebuild was configured for an outerworks shot that relied heavily on the pixy for correcting for yaw misalignment.

The pixy was really easy to get set up. You can tune the pixy offboard the robot using pixymon. Once we had that all tuned, we were successfully tracking the target within an hour of mounting it to the robot.

[kylemikableh]

Originally Posted by Jeremy Germita

We love the pixy. Rev1 of our rebuild was configured for an outerworks shot that relied heavily on the pixy for correcting for yaw misalignment.

The pixy was really easy to get set up. You can tune the pixy offboard the robot using pixymon. Once we had that all tuned, we were successfully tracking the target within an hour of mounting it to the robot.

Wow, that looks amazing! How does this translate to robot code? Must had taken hours to accomplish that.

[Citrus Dad]

Your offseason robot performed amazingly well at Chezy Champs! Congrats!

[Jeremy Germita]

Originally Posted by kylemikableh

Wow, that looks amazing! How does this translate to robot code? Must had taken hours to accomplish that.

After tuning the pixy's settings to only give us data on the target's X position through the analog out, we used a simple P-controller on the drivetrain. It could have been tuned a little better, but we found that in outerworks shot config, our drivers were good at getting close enough for the P loop to get us into position reliably. You can take a peek at the code here, although it isn't anything groundbreaking - lines 84-89.

This never ended up being used on the competition field at Chezy Champs due to us pivoting to a ramp camp shot at the base of the batter that was a lot easier and faster for our drivers to hit. It was a pesky measurement error in translating rev 1's shooter exit angle to CAD for rev 2 that wasn't caught until after we had cut the plates for rev 2. Measurement was 30 deg from vertical, CAD was 30 deg from horizontal! Ended up working out in the end, but expect to see one more big revision for Battle at the Border in a few weeks...