Signal-Driven Swarm Intelligence and Evolutionary Robotics
Dr Olaf Witkowski from MTI Ltd in Tokyo, will present the Department of Computer Science Colloquium, with a talk entitled, "Signal-driven Swarm Intelligence and evolutionary robotics".
Colonies of bacteria, swarms of insects and flocks of birds all exhibit a swarming behavior based on local interaction and individual decision making. Identifying the minimal features leading to the emergence of such behavior is fundamental to our understanding of collective behavior. Decentralized swarm control is crucial in fields of application where
control is costly, such as space exploration or nano-robotics.
In this talk, Olaf will introduce a minimalist model of swarming based uniquely on agents exchanging basic signals between each other. The agents evolve a swarming behavior through an asynchronous evolutionary algorithm,
surprisingly allowing them to reach goal areas which they are unable to detect directly.
This minimalist approach paves the way for future research on the emergence of signal-based swarming as an efficient collective strategy for uninformed search. In the long run, the hope is to implement swarms of decentralized robots capable to reach higher levels of problem solving.
Olaf Witkowski graduated with an MSc in Civil Engineer in 2008 from the University of Louvain (UCL) in Belgium, and was a co-founder and Chief Research Officer at Commentag LLC in Belgium, designing the first semantic search engine for Twitter. He obtained his PhD in Computer Science from the University of Tokyo, Japan in 2014. Olaf is currently a researcher at MTI Ltd in Tokyo, and a post-doctoral research fellow at the University of Tokyo. His main goal of research is to understand the mathematical foundations of cooperation and collective behavior, the evolution of cognition and the origins of life.