Dynamic data-driven decisions

Self-aware autonomous vehicles

Using both physics-based and data-based models to issue better decisions than by just using data alone.

Learn more ↓

Oden Institute Willcox Research Group

On-board dynamic decision-making

A self-aware aerospace vehicle should be able to dynamically adapt the way it performs missions by gathering information about itself and its surroundings and responding intelligently.

The challenge is to achieve the tasks of sensing data, inferring the vehicle state, updating the vehicle flight envelope, and replanning the mission in real time — executing online models and exploiting dynamic data streams — while also accounting for uncertainty.

We use component-based reduced models derived from offline physics-based simulations that model the wing's aerostructural response. We update these reduced models in real-time using dynamic sensor data together with rapid online classification methods.

Wing strain sensor measurements and flight loads data

Structural model of wing stress and strain under different damage conditions

Data-to-decisions model that supports UAV dynamic mission re-planning in response to in-flight structural damage