A comprehensive simulation platform for developing, testing, and validating autonomous systems.
Unreal Engine 5 with hardware ray tracing. Physically-based materials, dynamic weather, and time-of-day simulation.
PhysX-based rigid body dynamics with custom aerodynamic, hydrodynamic, and ground contact models running at 1000 Hz.
High-fidelity LiDAR, camera, radar, IMU, and GPS simulation with configurable noise models and failure injection.
Scalable multi-agent simulation with inter-agent communication, coordination, and emergent behavior modeling.
Native Gymnasium and ROS 2 APIs for training RL policies with domain randomization and curriculum learning.
Connect real controllers, autopilots, and compute modules to the simulation loop for integrated system testing.
Drones, UAVs, eVTOL
Autonomous vehicles, UGVs
USVs, AUVs, submarines
Any rigid body platform
$ autonomysim status
✓ AutonomySim — Running
Engine: Unreal Engine 5 | Physics: PhysX
World: san_francisco_downtown | Agents: 4
─────────────────────────────────────────────
Sensors: LiDAR-128ch, Camera×6, Radar, IMU, GPS
Physics Rate: 1000 Hz | Render: 60 FPS (RT)
RL Policy: nervosys/av-policy-v3 (loaded)
─────────────────────────────────────────────
Vehicle Types: UAV | UGV | USV | AUV
Environments: 12 loaded | 48 available
Telemetry: streaming at 50 Hz