Autonomous Machine for Inspecting Gas & Operations (AMIGO)

Summary

As a part of the AMIGO project, I...

➤ developed an anomaly searching control algorithm,
➤ trained custom RL policies in Isaac Lab,
➤ wrote a ROS2 based RL deployment framework,
➤ wrote additional ROS2 functionalities (actions, services), and
➤ supported our rapid prototyping efforts.

The goal was (and is) to develop an autonomous robotic platform for powerplant diagnostics. We outfitted the Unitree Go2 EDU quadruped with ...

➤ Jetson AGX Orin,
➤ 2D LiDAR,
➤ ZED X camera,
➤ RealSense camera,
➤ Raspberry Pi, and
➤ gas sensors.

With our additions, the robot could autonomously navigate indoors, leveraging our custom Nav2 implementation amigo_ros2. The team at EPPL are currently adding outdoor (GPS) navigation functionality.

RL Locomotion Training and Deployment

Powerplants often exhibit difficult terrain for a quadruped to traverse. Specifically, most plants are littered with open-backed staircases that pose a great challenge for blind locomotion policies and controllers. Too often, the robot (especially a smaller quadruped like the Go2) steps into the open region and falls face first into the stairs.

To tackle this issue, I acquainted myself with RL methods for robot locomotion. Starting from the basic Isaac Lab Go2 environment, I edited the reward structure, added privileged observations (for asymmetric Actor Critic NNs), implemented increased domain randomization, and abused my gaming laptop's GPU. To my surprise, once I had a policy that looked ~acceptable~ in the sim, I couldn't find a deployment framework for the Go2. So, I built my own and called it go2_rl_ws.

Attributions

Funded by Steam Solutions, and housed by the Engineering Physics Propulsion Lab @ ERAU.