Mondrian Madness and the ROMI's

Congratulations to the NE teams that were the winners in the ROMI challenge MONDRIAN MADNESS. While our team didn’t place, I was still proud of my students for learning the RAMSETE COMMAND that they can apply to our FRC machines! I was fascinated by the low times some of your teams came up with, so I am curious if anyone has any tips to share to make the ROMI’s more accurate?


We were also very impressed by the winners of Mondrian Madness. (Congratulations!!!) We are still trying to replicate the sub-4-second run times here. Even knowing that it is possible, it is an elusive goal.

Make sure your Romi image is fully updated. The WPILib team recently released a couple updates to the Web Service that apparently addressed a couple of gyro issues. Also remember to recalibrate your gyro after you update.

We found that the Romi was much more reliable at following trajectories when we reduced its top speed, but that worked against faster run times. I think getting good characterization constants for your Romi is crucial. Also, if your team has more than one Romi you may find that some have more stable IMU performance than others.

My only other observation is that mapping a smooth trajectory is important. Our Romi lost speed during sharp corners, so making turns smooth helped cut runtimes a bit.

Our best submitted times were about 2X the winning runs, so I am curious to hear some tips from the top performers.

I had a question about this as we move forward. Do people think batteries play a role in speed or handling? Specifically, do you think alkaline batteries versus NIMH with the higher voltage will equal more speed? We have not had time to test as we focussed most of our energy on the Game Design challenge for the past month.

This is certainly possible that battery voltage and charge state could play a role.

I also expect that differences between assembly play a role. I found one of our ROMIs had the encoder magnet pressed at an improper distance. It had a large impact on that side of the drive. Similarly, a different robot had an issue with wheel press-fit distance that caused issues with driving.

1 Like

Our team has only the one ROMI, but may get more. We do have the latest update, but we found that the ROMI never did the same thing twice, so it might be interesting to see if some are more stable than others.
I think we need to learn a bit about smoothing out the corners. Our team just put in coordinates at certain points along the path, but did not consider a spline. Perhaps we should have!

This topic was automatically closed 365 days after the last reply. New replies are no longer allowed.