9:52 / 55:39
https://s-rl-toolbox.readthedocs.io/en/latest/index.html
I came across this toolbox because with the robot, I am not sure what’s wrong, saving and loading policy files for ARS.
I asked about it here: https://pybullet.org/Bullet/phpBB3/viewtopic.php?f=24&t=13005
So I was considering using some “stable baseline” RL algorithms. They have an implementation of PPO, which is another recent algorithm.
Checked out https://github.com/agilityrobotics/cassie-gazebo-sim
There were some extra steps, as per usual. Ubuntu 18.04? https://automaticaddison.com/how-to-launch-gazebo-in-ubuntu/ – We need to change the URL to url: https://api.ignitionrobotics.org
cd ~/.ignition/fuel/config.yaml
and I needed to set some envs
GAZEBO_PLUGIN_PATH=/opt/cassie-gazebo-sim/plugin/build
GAZEBO_MODEL_PATH=/.gazebo/models
/.gazebo/models/cassie# gazebo cassie.world
It loads a derpy cassie robot
Then /opt/cassie-gazebo-sim/plugin/build# ./cassiectrl
runs the sim, which doesn’t do anything.
But the https://github.com/agilityrobotics/cassie-gazebo-sim/tree/master/plugin/include code is interesting, for remote controlling code in C using UDP. UDP is a good idea for remote control. Sends structs. Very straightforward. Nice. ZMQ probably nicer though.
Looks like it integrates with a fancy https://www.elmomc.com/ motion control company thing. Nice big UI. But yeah. Cassie robot is much too complicated.