Control Scheme for Distributed Computing in Robot Networks Destined to Humanitarian Demining

Abstract: 
Multi-robot systems are projected as novel tools to help humans in hazardous activities such as humanitarian demining. Nevertheless, controlling a robot group behaviors is not a trivial task, since they need to process information suitably so those team goals are achieved in a systematic form. Previous work has not made a clear distinction between the role played by coordination and cooperation, or the importance of conformity in robot networks. In this paper, a distributed control scheme is presented wich considers consent variables. The variables values are set by the robots in a consensus way to coordinate common actions. Finally, robots cooperate each other using local interactions sharing system information. Bottom-Up methodology is used to develop the distributed control strategy, which has been validated using simulation and experimentation with real robots of the Robotarium platform, an accessible mobile robot testbeds. The results show that it is possible constructing collective behaviors using a control scheme that organize the way in which the information is computed by robots in a network system. In particular, the Progressive Reduction behavior, developed for the Area Reduction task, show how robots can explore suspected areas of contamination with antipersonnel mines to detect them.
Author(s): 
Victor Sillerico-Justo
Date: 
August, 2018