As of now, 2 apriltag alignment strategies that I’m aware of involve creating a trajectory on the fly to some setpoint near an apriltag and having the robot follow it, and using a pid speed controller to drive the robot based on x, y, and theta offsets from some apriltag pose. I’ve seen both systems work, but I don’t know which one is better. Has there been one that your team has found to be faster and more consistent?
2 Likes
I have been using pose estimation to get our robots field relative pose. Then I use pathplanner’s Pathfinder feature to avoid obstacles and generate a trajectory on the fly to anywhere in the field as well as aligning with apriltags. For pose estimation you might need to play around with tuning standard deviations. It’s been working pretty well for me.