Correct-by-Construction Controller Synthesis using Gaussian Process Transfer Learning

This project explores improvements to embedded control software for safety-critical cyber-physical systems with applications in autonomous transportation, traffic networks, power networks, and aerospace systems. These systems often have complex dynamics that are difficult to obtain in a closed-form and ensure their safety.

This project investigates a novel correct-by-construction controller synthesis scheme for these systems by embracing ideas from Gaussian processes and control Barrier functions. If successful, this could allow safety controllers developed for one type of autonomous vehicle to be transferred to another of a wholly new type – or for use in a new environment all together – while still ensuring the original safety guarantee. This would enable safe deployment of multiple (similar) systems using the same control architecture specifically designed for only one of them; hence, a significant increase in efficiency in the design process. The resulting algorithms of this project will be tested on underwater and aerial vehicles with an eye to future applications to other CPS domains.