Siobhan Duncan

Swarm robotics has the potential to revolutionize the robotics world, as swarm systems can tackle a wide range of applications more effectively and efficiently than any single robot. In swarms, no single agent, or group of agents, is in control. Nor is there a central knowledge or master overview of the swarm's state, it's goals, priorities, or the general environment it's located in. Stigmergy is a key factor in the coordination of swarm systems as it enables the emergence of global intelligence through the interactions agents have with stimuli found in the environment. Agents leave traces of their actions within a medium, such as the environment, which then increase the probability to prompt new actions when sensed by other agents. A famous example are the pheromone trails left by ants whilst foraging. To date, stigmergy has mostly been confined to simulation-based or table-top experiments. However there exist a few exceptions such as using RFID tags as the medium for stigmergy, or tying the pheromones to the robots themselves. Contrary to past efforts, which have tried to mimic biological mechanisms and found those hard to realise in hardware implementation, my research focus is to leverage modern scalable technologies, such as cloud computing, the Internet of Things (IoT) and embedded distributed devices, to bring stigmergy out of the laboratory and robot swarms into the real world. My current research hypothesis is thus: "It is possible to extract general principles subtending stigmergy-based swarm systems, and use modern technologies, such as GPS, IoT and Cloud computing, to apply them to real-world robot swarms, in order to improve their scalability, flexibility and robustness."