Online Multi-Robot Exploration with Communication Restrictions

Abstract

In the aftermath of a disaster, such as an earthquake, it can be hours or even days before human rescue teams can safely enter damaged structures to search for and extract survivors. This critical time can be productively utilized to provide advance information to the search and rescue efforts by deploying robots to explore the disaster environment before it is safe for the human response team to enter. This permits the response team to develop a more definitive and efficient plan of action for rescuing survivors or shoring up the area to prevent further destruction. However, there are limitations on what robots can accomplish in disaster environments, with communication significantly restricted and the environment essentially unknown. To overcome these obstacles, this research has focused on creating novel algorithms for online, multi-robot exploration to achieve full coverage of an unknown environment using small, basic robots. The primary contributions in this thesis are two distributed algorithms for exploration using small teams of robots with limited communication. The innovation in these algorithms comes from how the robots disperse into and subsequently explore the environment, even with communication restrictions. We provide theoretical analysis that shows the algorithms will achieve full coverage of the environment, and return all functioning robots to the entry point. The more restricted algorithm uses the minimum number of message types. We demonstrate the algorithms' functionality using simulations and experiments using physical robots, and compare their performance with other algorithms. The primary algorithm is able to perform on par with other algorithms in spite of the communication restrictions, in multiple environments.