Robotics Motion-Control Loop with Safety Layer

Reference policy + low-level controller + robot dynamics + safety filter, with stability indicator.

Prompt

Create a robotics motion-control loop diagram with an explicit safety layer.

Layout:
- Show a closed-loop control chain: reference trajectory -> high-level policy -> low-level controller -> robot dynamics -> sensors -> state estimator -> back to policy and controller.
- Insert a safety filter between controller and robot dynamics.
- Add a stability / constraint monitor connected to the safety filter.
- Show disturbance input entering the robot dynamics block.
- Add a small inset showing safe set boundary and current state trajectory.

Style:
- Academic control-systems diagram on white background.
- Navy blocks, teal feedback arrows, coral safety / constraint path, amber disturbance arrow.
- Use crisp orthogonal connectors and readable math-style labels.
- Keep text short and aligned.
- Suitable for robotics, reinforcement learning, and safety-control papers.

Use in Generator

When to use

For robotics, control, and embodied-AI papers introducing safety wrappers around learned policies.

Variations

Without learned policy

Replace the learned policy with a classical motion planner (RRT* + trajectory optimisation). Keep the safety filter and explain that the filter now serves as a runtime certificate.

Tips

Always include the safety filter as a separate block. Embedding it inside the controller hides the safety story.
Show the instability score over time. Numerical safety arguments need a visible signal.
Label sensor noise on the observer. Real systems do not get clean states.

FAQ

Can I show fallback behaviour?

Add a small inset showing the policy output and the filtered reference diverging at the moment R exceeds the threshold, with a label "fallback engaged".