Heiko Hamann (Science of Intelligence), “Group performance and scalability in collective systems”

Abstract:
Scalability can be challenging in groups of collaborating agents, such as animals, robots, or computers. While a small group may work efficiently together, a bigger group may be slowed down due to increased needs to communicate and synchronize or due to other scarce shared resources. We go through a number of examples for observed system performance over system size and find common features. Based on these findings, we define a simple mathematical model that catches these main features and can generically be applied to different domains, such as robotics, computing, and sensor networks or possibly even human groups. We end by speculating a bit of what might be common to all of these systems and what might be the underlying drivers for the limits of scalability.

The Zoom Link will be sent the day before the lecture. (Contact communication@scioi.de for specific questions)