Communication Structures in Underwater Swarm Robotics Systems

A swarm of robots has unique characteristics that allow it to handle complex tasks efficiently. In this work, the deployment of a swarm of robots in a body of water is studied using computer simulations to track and identify sources of plume pollution. The water body is modeled as a section of a lake. The pollution source is a marine plume specified at a location unknown to the robots, which the robotic swarm is tasked to track efficiently. Heterogenous swarms of robots are employed which allows exploration of various behavioral archetypes of the robots as well as the distribution of archetypes among the swarm to simulate a diverse body of robots, and this is compared to homogeneous swarms of robots too. A computer simulation-based study is carried out to investigate the speed and accuracy of marine pollution plume detection by swarms with varying extents of heterogeneity. It is demonstrated that the diversity of robots can be beneficial to a swarm of robots if the archetypes are individually productive but could be harmful otherwise. Different extents of diversity are useful depending on the archetypes present in the swarm.