BAXTER TUTORIAL TRAJECTORIES

The safe pipeline is Record -> Visualize -> Playback, this verifies your trajectories before going and trainig your model to generate new ones.

Record

Execute the recorder in this baxter-python3 project, clone the repo if you don’t have it.

I suggest to do it from the pc, without ssh into robot and cloning stuff there, you can esily mess up stuff, also from a pc you can visualize immediately and copy easily the recordings files to your usb.

(If you ssh in the internal robot computer, it’s inside igor folder, remember to activate a ros environment before, es source ~/alper_workspace/activate_env.sh )

Check the options: You can specify the arm(s), recording rate, fixed duration, recordings name, and if you want also the joint-space recorded.

python example_trajectories_recorder.py --help

Visualize

You should now have the trajectories in .CSV.

Execute the visualizer inside the same folder, it has also some cool options to check.

(If you ssh in the internal robot computer, you can’t visualize there the trajectories, once copied them in your pc, use the script from there)

It will show all cartesian trajectories.

python visualizer_cartesian_trajectories.py

Playback

Execute the playback code inside the same baxter-python3 interface folder, test it with the trajectories recorded and use this to play the new trajectories that you generated.

(If you ssh in the internal robot computer, it’s inside igor folder, remember to activate a ros environment before, es source ~/alper_workspace/activate_env.sh )

python example_cartesian_space_playback.py --help