This repo contains the configuration files corresponding to the hyperplan repo
results directory stores the results of running the hyperplanning optimization and then testing of the said configurations
The Only configuration files we will be running are:
- box_pick-*.yaml
- tall_pick-*.yaml
- tall_place-*.yaml
- thin_vert_place-*.yaml
We will only be running configuration files with the loss functions of:
- execution time
- path length
- memory_loss (when that is implemented)
NOTE: We have updated the accross_groups to only run on the small_pick configuration files
See the corresponding images of each dataset for a visualization of start and goal.
pathxxxx.yaml is a valid trajectory requestxxxx.yaml contains the start and goal configurations in joint_space scenexxxx.yaml contains the geometric description of the obstacles scene_sensedxxxx.yaml contains the octomap representation of the obstacles
The box_pick and table_under_pick contain problems that are in "open space" so I would assume similar parameters work best for them
The tall_pick, small_pick, thin_vert_pick are bookcases that are narrow
tall_place and thin_vert_place are similar to the "pick" versions but instead of the stow position the robot starts inside a random shelf
The thin_vert_place is the hardest one of all so far "especially the octomap version"