Achieving Connectivity between Wide Areas Through Self-Organising Robot Swarms using Embodied Evolution

Abstract

Abruptions to the communication infrastructuremight occur occasionally, where manual intervention by dedi-cated personnel is needed to fix the interruptions, restoring com-munication abilities. However, sometimes this can be dangerousto the personnel carrying out the task, which can be the case inenvironmental disasters like earthquakes or toxic spills or in theoccurrence of fire. Therefore, personal risks can be minimised ifautonomous robots are deployed to achieve the same outcome: toestablish a communication link between two previously connectedbut distant sites. In this paper, we investigate the deployment ofmobile ad hoc robots which relay traffic between two sites. Inorder to get the robots to locate themselves appropriately, wetake inspiration from self-organisation and emergence in artificiallife, where a common overall goal may be achieved if the correctlocal rules on the agents in the system are invoked. We integratethe aspect of connectivity between two sites into the multirobotsimulation platform known as JBotEvolver. The robot swarm iscomposed of Thymio II robots. In addition, we compare threeheuristics, of which one uses neuroevolution (evolution of neuralnetworks) to show how self-organisation and embodied evolutioncan be used within the integration. Our use of embodiment inrobotic controllers shows promising results and provides solidknowledge and guidelines for further investigations.