A parameterized model for supporting self-organization in multi-robot networks

Abstract

The availability of robust and power-efficient robotic devices boosts their use in a wide range of applications, most of them unfeasible in the recent past due to environmental restrictions or hazard to humans. Nowadays, robots can support or perform missions of search and rescue, exploration,surveillance and reconnaissance,or provide a communication infrastructure to clients when there is no network infrastructure available. In general,these applications require efficient and multi-objective teamwork. Our work in this project is to design and test, in different scenarios, a general model for distributed robotic control that considers different task objectives depending on the application domain. As a starting point, our research group has already developed techniques that combine distributed connectivity maintenance and failure mitigation for mobile robotic teams, two important features for domains that require establishing an ad hoc communication network whilst tackling possible robotic failures that could potentially disconnect the network and thus reduce collaborative task achievement. The proposed model will extend such techniques to a more general model that can be adapted to different domains, with simulation tests in a UAV stealth scenario and simulation and real settings in a robotic communication coverage scenario. Formal development of the parametrized model w.r.t. convergence and performance bounds will also be conducted. (AU)