Experimental Testbed for Large Multirobot Teams

摘要

Experimental validation is particularly important in multi-robot systems research. The differences between models and real-world conditions that may not be apparent in single robot experiments are amplified because of the large number of robots, interactions between robots, and the effects of asynchronous and distributed control, sensing, and actuation. Over the last two years, we have developed an experimental testbed to support research in multirobot systems with the goal of making it easy for users to model, design, benchmark, and validate algorithms. In this article, we describe our approach to the design of a large-scale multirobot system for the experimental verification and validation of a variety of distributed robotic applications in an indoor environment.