Swarm robotics: Designing scalable self-organized collective robotic platforms that combine the strengths of engineered and biological systems

Add to your list(s) Download to your calendar using vCal

• Dr Eliseo Ferrante (School of Computer Science, University of Birmingham)
• Monday 25 March 2019, 11:00-12:00
• Computer Science, The Sloman Lounge (UG).

If you have a question about this talk, please contact Hector Basevi.

Swarm robotics studies the design of collective behaviours for swarms of robots that, by interacting with each other using simple local sensing and communication, achieve a collective-level emergent behaviour through a process called self-organization. From an engineering perspective, the goal is to produce scalable collective behaviours able to tackle tasks in large, unstructured, and gps-denied environments, by developing models inspired by biological systems that live in groups, such as ants, bees, birds, fish. The focus is understanding and modeling the key the behavioural and interaction rules used by these systems to achieve self-organization, and apply this knowledge to design swarm robotics systems with minimal hardware and communication requirements.

In this talk I will summarize at a high level the main projects I have been working on. In the first part, I will discuss the idea of introducing a small proportion of robots that are more capable than the other members of the swarm. These robots act similarly to virtual leaders, and they are used to counterbalance the swarm tendency to self-organize to random collective behaviours as opposed to those needed for the task at hand. For example, I will present results I achieved in collective motion, where these more capable robots allow the swarm to react to changes in the environment and to deal with conflicting information about the task. I will also present recent results whereby the same idea is used to control a self-organized aggregation process with robots with minimal sensing capabilities.

In the second part of the talk, I will give an overview of the other research projects I have worked on. This will include: another project on collective motion with minimalistic sensing, which represents also a major contribution in the field of statistical physics; a project on the evolution of task specialization, which involved the development of a novel evolutionary computation technique to automatically synthetise a complex collective behaviours starting from basic building blocks; a series of research projects on collective decision making in robot swarms, which includes recent results achieved in the context of dynamic environments.

Videos: https://www.youtube.com/watch?v=teRFP3FshDE https://www.youtube.com/watch?v=ozwy26yvB1U https://www.youtube.com/watch?v=uEXK-b3EAHU&t=1s https://www.youtube.com/watch?v=M2nn1X9Xlps&t=2s

Host: Prof Hamid Dehghani

Biography: Dr. Eliseo Ferrante owns a Ph.D. in Engineering Sciences delivered by the Université Libre de Bruxelles (ULB – Belgium) in 2013, a MSc in Computer Science from the University of Illinois at Chicago, and a MSc in Computer Science Engineering from Politecnico di Milano (Italy). He is currently a Lecturer at the University of Birmingham, where he contributes to both Dubai and UK campuses. Dr. Ferrante has authored more than 30 peer-reviewed publications, among which 17 publications in international journals with peer review, and 24 articles between peer-reviewed conference, workshops, and video proceedings. Some of Dr. Ferrante journal articles have been published in journals with high impact factor, including Physical Review Letters (IF 2016: 8.46), IEEE Transactions in Cybernetics (IF 2016: 7.38). He also authored a survey article with more than 700 citations since 2013. His Google Scholar H-index is 18, and the total number of citations is over 2000. Dr. Ferrante’s research was featured on international and national magazines, including Science Magazine and IEEE Spectrum.

Dr. Ferrante’s research focuses on swarm robotics studies from an interdisciplinary perspective comprising computational, statistical physics, and evolutionary models of collective behaviors. Some of the phenomena he studies include collective motion, task specialization, and collective decision-making in artificial agents and robots. His methodological expertise includes computer simulations, real robot experiments, evolutionary and mathematical models.

Website: https://scholar.google.co.uk/citations?hl=en&user=cVjuZJoAAAAJ

This talk is part of the Artificial Intelligence and Natural Computation seminars series.

This talk is included in these lists:

• Artificial Intelligence and Natural Computation seminars
• Computer Science Departmental Series
• Computer Science Distinguished Seminars
• Computer Science, The Sloman Lounge (UG)
• computer sience

Note that ex-directory lists are not shown.