James Butterworth

Slider 1

Contact

Bio

I received my undergraduate degree in Artificial Intelligence from University of Liverpool in 2016. I went on to study a CDT with the Risk Institute for which I received my MRes in September 2017. I first discovered Genetic Algorithms and Neuroevolution during a second year undergraduate course when a group of colleagues and I wrote a simulation that evolved fighting agents. I was completely fascinated at how a computer program could learn such sophisticated control policies. Since then I have continued to work in this area in order to really discover how powerful these algorithms are.

My research interests are mainly in the area of applying machine learning techniques such as Neuroevolution to robotic swarms. Although, I am interested in anything to do with ML, Deep Learning, Multi-Agent Systems or Decision Making under Uncertainty

Research project title: Decentralised Mapping of Indoor Hazardous Environments Using MAVs

Project description: This project aims to develop a technical, algorithmic and mathematical framework that enables a swarm of MAVs to perform decentralised mapping of hazardous environments. The main issues to tackle are the areas of coverage, navigation, vision and map fusion algorithms as these all must be addressed for high quality surveillance and mapping system. This project will look at how to apply the ideas of Genetic Algorithms and Neuroevolution to evolve control policies for the agents in the swarm.

Supervisory team: Prof Karl Tuyls; Dr Paolo Paoletti

ORCID: 0000-0002-6446-6577