Department of Computer, Control and Management Engineering

Quorum sensing is a key mechanism enabling coordinated behaviour in populations of autonomous agents, and is extensively stud- ied in biological systems spanning from bacteria populations to social insects. In swarm robotics too, quorum sensing aroused much interest, but remained mostly constrained to the implementation of collective deci- sions. Here, we propose protocols to estimate the quorum level that are suitable for resource-constrained robots, and evaluate the precision and speed of the quorum assessment across a large spectrum of swarm state conditions. Through systematic experimentation, we evaluate the pro- posed protocols for different swarm densities and working area sizes. Our findings shed light on the trade-off between computational requirements and expected performance, aiding in the selection of appropriate quorum sensing protocols for future swarm robotics research.