Elliot Hogg

Email: eh17402@bristol.ac.uk

Supervisors: Arthur Richards and Sabine Hauert

PhD Project Title: Evolving Behaviour Trees for Supervisory Control of Robot Swarms.

Research Group: Swarm Robotics (Thales T-B Phase)

Biography

Principle Aim:

Apply evolutionary learning to determine effective swarm control strategies for real-world deployment.

Research Project:

Supervisory Control of swarms is essential to their deployment in real world scenarios to both monitor their operation and provide guidance. We aim to explore how humans might influence swarms to improve their performance, using different mechanisms with limited information and influence. Human trials have high time and logistical overhead, so we explore the use of artificial evolution to design supervising strategies, as a surrogate for a human supervisor. We use behaviour trees to represent human readable decision strategies which are produced through evolution. With this new approach, we aim to assess a large array of scenarios in which we would want to control a swarm and gain understanding of how to design these systems towards human control in the real world.