2 Hatches from Level 2
16 Likes
Very impressive
Thanks. We spent a lot of time messing with the overall robot path to be able to drop down from level 2 consistently, and we are really happy with the result (and the extra 3 points 
) .
Could you explain how you integrate vision with your paths smoothly? It’s interesting how the two PIDs don’t conflict against each other.
1 Like
Primarily, we use path following (based on pure-pursuit) to navigate in auto. This year, we wanted to be able to incorporate vision into our autos without the need to completely stop and align. To do that, we mark one of our points (out of our sequence of points that make up the path) as the point at which vision begins. Once the pure-pursuit controller thinks that it has reached that point, it will switch over to the same vision controller that we use during teleop to automatically drive to targets.
To know when we have arrived at the target, we wait 3 update cycles after the vision target is outside the limelight FOV. Our limelight is in a position that makes this a convenient solution.
This looks awesome, and so smooth! Excited to compete with you guys at SVR this weekend
This topic was automatically closed 365 days after the last reply. New replies are no longer allowed.