Department of Computer Science and Technology, West Cambridge

The Research Associate / Research Assistant will work together with a team of students and research collaborators on the development of control policies that facilitate the coordination of large-scale aerial drone systems. The ideal candidate will possess hands-on experience with designing and implementing reinforcement learning algorithms for robotic systems. They will be working with a team composed of PhD students, Undergraduates and Postdocs that is developing novel multi-robot architectures for practical, real-world settings. Current solutions to many of the coordination challenges use simplifying assumptions and simulated operational environments to achieve success. Our team is pushing these assumptions to their limits, with the aim to break the barrier to real-world and outdoor deployments. As a result, we need to design systems and algorithms that are robust, while also offering scalability to larger swarms and highly dynamic environments. The RA will focus primarily on the concept design, implementation and verification/testing of multi-agent RL algorithms, but will also be involved in field testing and deployments. Therefore, we require someone with an excellent programming background, as well as rapid implementation skills.

The Prorok Lab in the Dept. of Computer Science & Technology, has a variety of robotic platforms (aerial and ground), and boasts expertise in controlling and deploying them in practice, as well as in designing coordination strategies for them. Our recent work on RL and graph neural networks (GNNs) demonstrate some of our key research directions relevant for this position. A high degree of self-motivation, research-oriented thinking, and a drive for getting real robotic systems up and running are all essential traits for this position.

https://www.jobs.cam.ac.uk/job/36052/