Title: Three recent research works in swarm robotics
Abstract: In this talk, I briefly illustrate objectives and achievements of three experiments that I have recently conducted with the support of my PhD students. All three works fall into the research area of swarm robotics/swarm intelligence.
In the first study, we have developed a control system to allow a swarm of robots to cooperatively transport objects that due to their mass, cannot be moved by single robots. The controller is a dynamic neural network synthesised using evolutionary computation techniques. The study shows that the robots do not need to feel forces to coordinate their action in order to find a common direction of transport.
In the second study, we have looked at the process of aggregation in swarm of robots with “informed robots”. These are robots that have been instructed on which site to aggregate, among those available in the environment. Our study looks at the proportion of informed robots required to gather all the swarm on a selected aggregation site.
In the third study, we have tested a probabilistic adaptive defence mechanism to allow a swarm of agents to limit the disruptive effects of a communication manipulation attack implemented as modification of data in transit events. The results of our study show that the proposed mechanism can be largely effective in multiple context differing for the proportion of malicious agents carrying out the attack.
I am happy to provide further details/readings for each of these works and to discuss about potential collaborations with anyone interested in pursuing interdisciplinary research work in the area of bio-inspired robotics, swarm robotics, swarm intelligence.