Bid Prediction for Multi-Robot Exploration with Disrupted Communications

Abstract

Teams of autonomous mobile robots have potential to contribute to surveillance as well as search and rescue operations. Larger and more complex disaster scenarios with high operational tempo, such as when delay may mean further loss of life, may benefit from cooperative teams of many robots working together for efficient search operations. Unfortunately, these scenarios also exhibit communications disruptions which can limit the ability for distributed algorithms to coordinate the actions of a team of coordinating robots. This paper will present an approach to overcome these communications disruptions by predicting the bids of disconnected teammates in a distributed auction over a spatially partitioned set of exploration tasks.