Driving simulator: an autonomous test car overtakes another car along the path its model predictive controller planned, drawn in blue, with a dashboard of its speed, steering angle and acceleration.

Controls & AI

We model how your product behaves and develop its control software. Where useful, we build artificial intelligence (AI) models that can run on the device.

What it covers

We simulate how a system behaves and design software to control it. This can include model predictive control, which uses a model to plan the next action. We also train machine-learning models on sensor data and adapt them to run on the device.

Where we’ve used it

For Haier, we modelled the physics of a magnetorheological damper to find which parameters matter most for its damping force. Our morphing airfoil was built to collect the wind-tunnel data that will train a model.

In a driving simulator, our model predictive controller planned a car’s path, overtook slower traffic and stopped for traffic lights.

What you get

Models, simulation results and controllers you can test against, with the code and documentation to take them into production.