Multi-robot Trajectory Coordination for Complex Vehicle Combinations

Coordinating the motion of multiple automated vehicles in a delimited area - or, more generally, multiple moving objects in a shared workspace - is a standing challenge in robotics research. In this thesis, approaches for the coordination of multiple mobile robots are investigated and extended to enable automation of large vehicles such as self-driving trucks with trailers operated in logistics distribution centers. Many existing algorithms are limited to simple circular or car-like geometries, which leads to unsatisfactory results for larger vehicle combinations. In the algorithms presented in this thesis, the limitation to simple geometries is overcome by introducing spatiotemporal occupancy tracking that allows for precise coordination of complex vehicle combinations. Various algorithms for deriving coordinated trajectories are presented, evaluated and extended. As disturbances in the execution of coordinated motion plans cannot be excluded in real-world applications, techniques for handling such disturbances are elaborated and tested. All contributions are combined in a software framework that is ready for deployment to real-world applications.