Example: ML Planning

Warning

🚧 This example is under construction. The implementation of this example is not yet publicly available. It is currently being prepared for publication and will be released as soon as possible.

Background

The ml-planning example represents a step toward end-to-end (E2E) planning. It demonstrates how a learning-based planning component can be integrated alongside a classical planning pipeline, focusing on comparing and switching between different trajectory generation modes during a running simulation.

This example covers:

• parallel execution of a classical planner and an ML-based reference trajectory planner

• orchestration logic for switching between planning modes

• additional sensor inputs required by ML-based planning components