The dynamic pathplanner allows us to set different size/strength/shapes of obstacles to change specifically how the path is planned around the obstacle.
Right now we just make one kind of obstacle for anything detected.
We need to expand the obstacle detector camera wrapper to categorize obstacles differently based on what the camera says.
Potential use case: Avoid same-alliance robots more than opposite-alliance robots. Or vice-versa, if we trust our partners to get out of the way faster.
Part of this may also require photonvision to have two pipelines on the same camera (one that detects red blobs, one that detects blue). Alternatively, use the ML detected object category idx to differentiate red/blue
The dynamic pathplanner allows us to set different size/strength/shapes of obstacles to change specifically how the path is planned around the obstacle.
Right now we just make one kind of obstacle for anything detected.
We need to expand the obstacle detector camera wrapper to categorize obstacles differently based on what the camera says.
Potential use case: Avoid same-alliance robots more than opposite-alliance robots. Or vice-versa, if we trust our partners to get out of the way faster.
Part of this may also require photonvision to have two pipelines on the same camera (one that detects red blobs, one that detects blue). Alternatively, use the ML detected object category idx to differentiate red/blue