All products
Image pending
Simulation
AI Motion Simulator
Motion cueing that learns the driver, not just the road.
A closed-loop motion platform whose washout filters are tuned by a reinforcement-learning agent rather than by hand. It observes vestibular mismatch across a session and adapts cueing gains per operator, cutting the false-cue rate that causes simulator sickness.
Specifications
- Degrees of freedom
- 6
- Payload
- 450 kg
- Control loop
- 1 kHz
- Cueing model
- Adaptive RL washout
- Latency, motion-to-photon
- < 18 ms
- Interfaces
- CAN FD, EtherCAT, ROS 2
Indicative figures. Final specification is confirmed against the application at quotation.
Features
- Per-operator cueing profiles learned in-session
- Vestibular mismatch telemetry exported to CSV and ROS 2 bags
- Hardware-in-the-loop bridge for ECU validation
- Deterministic 1 kHz real-time control core