Section: Computer Science and Engineering
Tutor: PERNICI BARBARA Major Research topic
:Coordinated navigation strategies for teams of communication-constrained autonomous mobile robots
Advisor: AMIGONI FRANCESCOAbstract:
One of the most recent trends in Autonomous Robotics involves the study of cooperative multi-robot missions in communication-restricted environments. In such settings, guaranteeing the ability to communicate between robots and/or between robots and human operators may be of primary importance for several reasons. For instance, in multirobot exploration tasks, communication between robots is fundamental for organizing a smart division of the environment to explore. In search and rescue scenarios, human operators should be continuously updated with video streams to be able to promptly detect potential victims. A last example is given by patrolling settings, where robots could need to periodically report their status and that of the inspected locations to a mission control center, whose deployed infrastructure is not able to cover all the regions of the environment. The communication issues in missions involving teams of autonomous mobile robots have been not fully addressed in the last two decades, and it is only in the last few years that some stable results appeared. The main reason is that, usually, much more constrained problems arise upon existing ones, so that in some cases even defining a proper performance metric to assess the mission quality in terms of communication becomes itself a non-trivial task. Most of the works presented in the literature deal with the communication issues by resorting to heuristic algorithms which, although proved to be effective (mostly in simulation), either rely on unrealistic communication models (e.g., the static circle, which may fail to model the presence of walls in indoor environments), or do not offer any bound on the quality of the obtained solutions. We feel that a deeper understanding of what can be accomplished by robotic teams involved in missions where communication requirements are of primary concern could have a strong impact on several real-world applications. Accordingly, the objective of this research is to contribute in building a methodology for developing effective decision strategies for the coordination of teams of cooperative robots subject to communication requirements, with a particular focus on the domains of search, exploration, and patrolling. For each of the studied coordination problems, we intend to proceed according to the following points: (i) give a precise problem definition taking into account one or more kinds of communication requirements, (ii) provide a rigorous study of the problem complexity, (iii) devise both exact and heuristic algorithms (whenever possible, with performance guarantees), and (iv) test the performance of our strategies on teams of simulated and real robots.