Forecasting and Reactive Intelligent Gliding for Autonomy via Thermal Exploitation (FRIGATE)

This project focuses on autonomous gliding in uncertain atmospheric environments, with a particular emphasis on exploiting thermal energy to extend range and improve mission efficiency. The core challenge is that thermals are transient and difficult to predict precisely, so an autonomous glider must do more than follow a fixed path. It needs to continuously infer where usable lift may exist and adapt its behavior online while still satisfying vehicle and mission constraints.

At a high level, the project develops an onboard path-planning framework that combines forecasting with reactive decision-making. The methodology centers on integrating planning with estimation, prediction, and control so that the aircraft can update its trajectory in real time based on both the predicted environment and newly observed conditions. In that sense, the project is about building an autonomy stack that can reason about energy, environment, and safety simultaneously rather than treating guidance and control as separate problems.