Autonomous Robot Work Cell Exploration Using Multisensory Eye-in-Hand Systems
The thesis presents a novel sensor-based approach to robotic exploration of partly unknown environments. Designed to facilitate automated work processes in flexible work cells, an efficient and reliable task-dependent exploration is performed by integrating flexible sensor systems, probabilistic environment representations, view and motion planning, and the fusion of information. Safe motion planning is achieved by multisensory data acquisition. The methods and sensors are evaluated in 3-D simulations resulting into design criteria for a multi-purpose exploration sensor, the 3D-Modeller. Experiments for exploration of work space, regions of interest, and combined missions are successfully performed. The developed method, which considers environment uncertainty in the planning process, enables flexible information gain-driven missions.