Flexible mission planning and dynamic reactive autonomy for heterogeneous unmanned vehicles operating in teams

Creating cooperative and intelligent behavior in a team of heterogeneous unmanned vehicles trying to achieve a mutual goal is a challenging task. The focus of this paper is on our technological approach for creating a flexible framework allowing individual agents organized as a team to conduct independent actions while still creating synergy effects. Translating a task into generic mission descriptions is one of the most central problems in this approach to create autonomous multi-robot systems. In addition to complying with operational and technical constraints, the resulting mission data needs to allow the handling of unforeseen events. We solved this by enabling each agent to make local decisions within global mission parameters. A system demonstrator that transforms the abstract tasks into a dynamic virtual environment in which the agent can react to the local situation based on virtual measurements has been developed and successfully tested in simulations as well as validated in specific real-world trials.