Trustworthy Evolving Functionality

The UKRI Trustworthy Autonomous Systems (TAS) Node in Functionality involves a team of social scientists, ethicists, computer scientists and engineers. We are exploring methods for designing autonomous systems (such as robots) with the ability to adapt how they function in response to changes in the world around them. This ability to evolve in functionality and adapt to current conditions has the potential to enable autonomous systems to better deal with uncertain real-world environments. However, having autonomous systems that can change the way they work on a day-to-day basis challenges our conventional methods for specifying, designing, verifying, and monitoring systems.  These types of adaptive systems also raise many questions around issues such as trust, ethics and regulation.

In this research programme we are developing a Design-for-Trustworthiness framework for adaptive autonomous systems. This considers the technical aspects of autonomous system design and also the broader contexts under which they need to operate.

We are developing three different technologies, with very different methods of adaptation, and using these to focus the development of our methods for ensuring autonomous systems are safe, reliable, resilient, ethical and trustworthy.