Knowledge-Based Framework for Human-Robots Collaborative Context Awareness in USAR Missions

Abstract

Urban search and rescue (USAR) missions can benefit a great deal from teams of mobile robots endowed with advanced perception capabilities. To effectively collaborate with humans, these robots should have situation awareness about their robotic and human teammates, for intuitive decision making. Moreover, robots should be able to contextually share information so that humans can benefit from augmented situation awareness provided by robots, and at the same time, actions taken by the robots be transparent to humans. In this paper, a knowledge-based framework for humanrobots collaborative context awareness in USAR missions is proposed. The main contributions are: an ontological representation of contexts at mission, agent, scenario, and team levels of the mission, a knowledge base integrating different tools required for such scenario, and an efficient and robust knowledge sharing strategy. The framework is efficient in terms of communication delay, capable to cope with communication failures and different event frequencies, and scalable in terms of team size.