Mission Management for Mobile Robots

For at least a decade a forthcoming advent of mobile robots in our everyday environments has been predicted by both media and researchers. Still we rarely meet relatives to the SciFi robots R2D2 or C3PO in our livingrooms or in our oces at work. Despite the constant evolution of advanced algorithms for mobile robot navigation, applications for mobile robots are eectively limited by the capabilities of planning and error-handling in the robot control software frameworks. To reach new goals in the application of mobile robots, it is necessary to design software frameworks, that are capable of planning and executing general navigation missions for mobile robots. Hereby can researchers and developers concentrate their eorts on developing navigation algorithms and robot applications instead of ad-hoc implementations of mission control. This thesis presents the analysis, design and implementation of a system for Mission Management for Mobile Robots. The system is a general framework, that allows application developers to design maps and navigation behaviors without programming in advanced software frameworks and operators to execute missions without the need of re-programming every mission. The Mission Manager uses a hierarchial graphbased world model, that allows fast topological world modeling, but additionally oers the possibility for true geographic representation, enhanced scalability and ecient planning through graph theory. Mission Execution is controlled through a separated deliberation and execution system where navigation behaviors are developed using a script-based generalized behavior model. The execution system implements a code generation model, that allows hard real-time error handling in low level execution and rule based feedback at behavior and planning level. The Mission Manager is a universal system that manages planning and execution of navigation mission for mobile robots and hereby enabling researchers and developers to dedicate their energy on maturing applications for mobile robots for our homes.