FLL bot inconsistancies over time

[TakeControlSolutions]

My FLL team has noticed our bots become inconsistent over time. We started the season with 2 identical bots, one primary with the most use and a secondary with very little use. After a couple of months of heavy use our primary bot became very inconsistent in a very short amount of time. We notice distances and turn angles not being as accurate as before as well as some jitter when going from a turn to a straight. For reference, missions that would run with a 90+% accuracy before, now run at a 10% or less accuracy.

We swapped the hub between the bots hoping that would be a fix, but that seemed to do nothing. Since we were close to competition, we just swapped our hubs back and went with the secondary bot. Things worked just as programmed once we switched to the secondary bot.

Now we are running into the same issue with our secondary bot after it has received more use. I'm having trouble distinguishing if it is a calibration issue or failing hardware (hub / motors). I believe there is a calibration for Pybricks, but I'm having trouble finding it, if someone could point me in the right direction, I would appreciate it. If it is hardware, what do you guys use to test for hardware accuracy?

[dlech]

The most common thing that I have seen that contributes to change in behavoir over time (with the exact same program and hardware) is dirt building up on the tires. This causes them to slip more easliy and the robot loses accuracy.

[laurensvalk]

Adding to that --- the mat too.

But if you find that accuracy depends on factors like these, you could perhaps think of ways to improve the design to make it less susceptible to such factors. You wouldn't want a dusty or sticky competition mat to throw off your robot on competition day.

For example, you could use the gyro for turns, use sensors to measure locations, or have the robot align itself against a wall.

[TakeControlSolutions]

We are cleaning the tires both with a lint roller and damp paper towel. That doesn't seem to be the culprit. We saw more slip last year with a smaller bot, but switched to a heavier bot this year.

We are using gyro on all missions, but it doesn't seem to be fully correcting for turns. After some more testing and messing with drive base settings (wheel base distance), it seems like left and right turns skew in different directions. A 90 degree right turn will work better with a 114 wheel base and a left turn works better a 102 wheel base. The actual measured wheel base is 104. Correct me if I'm wrong, but shouldn't gyro correct for this either way as it should be relying on the sensor to know when it has completed 90 degrees of turn?

The other thing that has me puzzled is that the failure seems to happen in a short amount of time and once it happens, we can't seem to get it to work right ever again. In our last meeting the bot went from working great at the beginning to failing less than an hour later. While middle school boys are not exactly gentile with stuff, there were no major drops during this time. This is the second bot to do the same exact thing.

At the meeting, we did have one more spare medium motor to swap out but not two. Things got better but not enough to call it good. Do these motors just need to be replaced every so often after heavy use?

[laurensvalk]

If it happened within one session, could the battery be a factor? If it is a heavy robot that runs on the edge of that it can do, maybe it only does it when fully charged?

How is the weight distribution of the robot? Could one support wheel have more friction than the other?

[TakeControlSolutions]

We charged the battery and experienced the same results. Battery charge doesn't seem to be a factor.

The bot is very symetrical and has not changed over the season. Attachments do cause weight shifts per mission, but all missions experience the issue regardless of weight distribution.

I am reaching out to Lego as well to see if they will replace motors. I'm pretty convinced the motors develop jitter over time that causes them to do an extra shift at the beginning of each movement.