Consensus in robot swarms
21st October 2022, 3:30 pm – 4:30 pm
Fry Building, 2.04
Robot swarms consist of large numbers of small robots cooperating to perform some task. Each robot has limited communication and sensing capabilities and the goal is to achieve coordination without a leader. One stylised problem studied in swarm robotics is that of determining the best option out of several, e.g., determining the best location for some purpose. A biologically inspired algorithm for this problem involves individual scouts visiting different locations, measuring their quality, and signalling their preference for the site for a period of time related to the quality. We present an analysis of this algorithm both for noiseless and noisy measurements of site quality.