Resource Efficient Path Planning for Mobile Manipulators in Recurring Machine Tending Tasks

Mobile Manipulators are on the path to becoming a general purpose automation solution with applications in fields of manufacturing, healthcare among others. However, the potential of mobile manipulators for high mix - low volume production is not exhausted, with operation to be expected more time-efficient, resilient and most of all more flexible. This paper presents an approach to mobile manipulator path planning for servicing tasks in a factory environment. The proposed algorithm enables time-efficient manipulator motion planning for recurring pick-and-place tasks through online path planning that is guided by offline-generated support data. This leverages the existing redundancy of information between recurring service tasks and reduces the overall computation needed. The results show that this approach is faster than sampling-based planning in static environments and produces comparably short trajectories in static environments with collisions created through human error.