Doctoral Project: Distributed Learning Control for Collective Problem Solving in Multi-Agent Systems
Part of research project: Dynamical Collective Adaptation & Learning
Description of the doctoral project
The successful candidate will use and extend methods from distributed and learning control theory to investigate the dynamics of collective learning. In particular, s/he will study (i) the interplay between prior knowledge, individual learning and knowledge transfer between agents, (ii) the relevance of the communication topology, and (iii) the effect of agent heterogeneity with respect to prior knowledge and learning strategies, dynamics and distortion of shared knowledge.
The following generic problem will be considered: A group of agents must explore an unknown environment, find objects, and pick them up. Picking up an object is a complex task that requires the agent to determine a suitable motor pattern. Agents learn these motor patterns by improving them over repeated trials. Between trials any agent can decide to share learned patterns or use patterns received from other agents.
Theoretical results and simulation results will be validated experimentally in a robotic testbed comprising several two-wheeled balancing robots with a mechanism that allows them to lift objects by performing a complex motion. The developed methods will be further validated, respectively improved, using observations from biological experiments realizing complex collective foraging problems for groups of fish. For this, the successful candidate will closely cooperate with a PhD researcher from behavioral biology.
Project start date: October 1, 2019 (an earlier starting date may be possible)
Applicants must hold a Master’s degree (or equivalent) in Engineering, Applied Mathematics, or a related subject. In particular, we require the applicant to have:
- a strong background in systems and control theory (previous experience in distributed control theory or learning control theory is an advantage)
- a genuine interest to pursue interdisciplinary research that involves control theory and both robotic and biological applications
- strong interdisciplinary communication and cooperation skills
- excellent English skills, both written and spoken