Our team has spent this season trying to get Path Planner working for our 2022 tank drive robot. Several weeks ago, we achieved trajectory following with events and managed to use the “reversed” flag to drive our robot backwards along the paths. However, since transitioning to our new tank drive robot, the paths have completely changed behavior, rotating out of control at the end.
Our odometry is consistent with the actual field and it doesn’t seem that tuning is the issue either. Both reversed and regular paths yield the expected final positions except for an unexplained rotation in the middle (see attachments). We suspect that there is an issue with how the pathplanner methods calculate the trajectories from the .path files, but there is not a lot to go off of.
Has anyone faced a similar issue and have any insights on what we may be doing wrong?
Have you updated your feedfoward constants for the new robot?
2 Likes
Yea we made sure to rerun SysID, but that turned out to not be the issue.
UPDATE: We ended up figuring out that, at some point, we chose to flip where the front of the robot would be, thus flipping the directions of the motors. While we’re still not sure why that may have completely messed up our auton, at least we’ve identified the cause.
Looks like you have it in holonomic mode. Since you have a tank drive, you should turn that off.
We’ve always had it turned off.
1 Like
Interesting. You should use the PathPlannerServer to monitor the status of the robot throughout the path and see what the odometry is vs where it’s supposed to be.
1 Like
Thank you for the suggestion, that’s a handy feature we never thought to use (we’ve been struggling with drawing the trajectories out on glass).
