Self-Optimizing Human-Robot Systems for Search and Rescue in Disaster Scenarios

摘要

The increasing capabilities of robot systems enable new fields of practical applica- tions for individual robots as well as multi-robot systems. But for some applica- tion scenarios like a fire or earthquake disaster current robots are still too limited to act fully autonomously in the disaster area. To overcome these limitations we consider a heterogeneous team of humans and robots complementing each other. Core application considered in this paper is a large burning warehouse with smoke making it difficult for fire fighters to search the building and to orientate them- selves inside the warehouse. Therefore, an assisting team of robots is surrounding the fire fighters searching the proximity, providing orientation data, and establish- ing a wireless communication infrastructure on a basis of a mobile ad-hoc net- work. The adaptation of the robots is achieved by applying principles of self- optimization on different levels of the human-robot system.\nIn this paper, we are considering self-optimization inside an individual robot to optimize its behaviour, within a group of robots, and in the entire system compris- ing of robots and humans. The focus of the optimization is the distribution of ro- bots by applying swarming behaviour for forming a mobile ad-hoc communica- tion network and performing map building.