Physical AI SIG

OSRA TGC Special Interest Group for Physical AI

This project is maintained by ros-physical-ai

Training & Execution Pipeline Working Group

Coordinator: Michael Gentner (BMW)

Overview

This group builds the end-to-end software workflow, enabling developers to take a trained policy and run it efficiently and robustly on a robot.

Scope

End-to-End Policy Deployment

Create a seamless flow to convert trained policies into performant ROS graphs, including critical steps like real-time data synchronization for inference.

Real-Time Execution & Performance

Develop and standardize core components, such as real-time capable executors, to ensure that deployed policies can meet strict performance guarantees.

Pipeline Tooling & Expansion

Improve the developer experience with simplified installation and testing tools while extending the pipeline to support advanced paradigms like Reinforcement Learning.

Roadmap

Long-term Vision

Roadmap content to be added

Current Priorities

Roadmap content to be added

Upcoming Milestones

Roadmap content to be added

Interested Members

Detailed member list to be added

How to Contribute

Contribution opportunities to be added