Motion planning and controls

Model any data required for simulation. Ingest HD maps, generate 3D worlds, use high or lower-fidelity models for physical or semantic sensors, and import physically accurate vehicle models.

Create complex scenarios with the GUI editor or generate them from natural language prompts. Test exact interactions or use machine learning (ML) behaviors to generate realistic situations. Run open- or closed-loop, object- or sensor-level, synthetic simulations, or log re-simulation. Run large test sets at scale in cloud simulation.

Analyze simulation results locally or in the cloud. Inspect stack, simulator, and pass/fail metrics in a user-friendly web interface or through Applied Intuition Copilot. Compare individual simulation results to inspect potential regressions.