ABSTRACT
Deployment of large teams of Micro Aerial Vehicles (MAVs) in real world (outdoor and indoor) environment independently to external positioning systems (such as motion capture or GPS) is a consequent step in current research in the field of autonomous flying robots. The aim of this talk is to present latest results of research of fully autonomous compact flocks of MAVs, which were achieved by Multi-robot Systems group http://mrs.felk.cvut.cz/ at Czech Technical University.
In particular, stabilization, control and motion planning techniques using Model Predictive Control (MPC) for steering swarms and formations of unmanned MAVs and heterogeneous teams of Unmanned Ground Vehicles (UGVs) acting together with MAVs will be presented. The proposed approach relies on onboard visual relative localization of individual MAVs, which is designed for utilization of multi-robot teams in dynamic environments without any need for global localization or motion capture systems. Such approach aspires to be an enabling technique for deployment of compact swarms of MAVs outside laboratories that are equipped with infrastructure for precise positioning of robots.
The swarming and formation driving behavior with the visual relative localization system in the control feedback will be shown in numerous outdoor and indoor experiments inspired by search and rescue, surveillance, and environment monitoring tasks. Finally, the latest results of transfer of the designed methodology into applications of distributed localization of RF transmission sources by an MAV formation and cooperative inspection and documentation of historical sites will be presented http://mrs.felk.cvut.cz/projects/cesnet.