Social-Aware Coordination of Multi-robot Systems Based on Institutions

Institutional robotics (IR) is an approach to the coordination of multirobot systems that draws inspiration from social sciences, namely from institutional economics. Using the concept of institution, it aims to provide a comprehensive strategy for specifying social interactions (e.g., norms, roles, hierarchies) among robots. In previous work, we have introduced a control methodology for multirobot systems that takes into account institutions in order to create an Institutional Agent Controller (IAC) that captures such social interactions. In this chapter, the IAC design methodology is validated in a case study concerned with a swarm of 40 real, resource-constrained robots which has to maintain wireless connectivity. We then investigate a second case study dealing with more complex social interactions, showing that institutional roles can effectively help a multi-robot system to coordinate and improve performance in a given task of social nature. Given the fact that institutions are one of the tools in use within human societies to shape social interactions, our intuition is that IR can also facilitate coordination with humans in scenarios involving many-to-many human-robot interactions. We discuss how the IR concepts and the IAC design methodology can be implemented in real-world scenarios where multiple robots must interact with multiple humans in a socially-aware manner. José N. Pereira and Alcherio Martinoli Distributed Intelligent Systems and Algorithms Laboratory EPFL, EPFL ENAC IIE DISAL, Station 2, CH-1015 Lausanne, Switzerland, e-mail: {jose.pereira,alcherio.