Vision-based path planning for flexible link robotics

In this thesis, vision aided flexible link robot positioning using the Camera Space Manipulation (CSM) method is developed. The primary motivation for this work is unmanned planet surface operations, specifically upcoming Moon and Mars missions. The earth-bound operator of the telerobot will command the manipulator to perform the required tasks. Typical tasks included in these missions will be digging, sample acquisition and construction. The robot's weight constitutes a big concern during shipment of the robots for planetary research. Using lightweight flexible robots will facilitate the launching. However, due to flexibility, rigid-body control strategies could not be applied to such structures. Off-line system commanding necessitates a vision aided control method. The communication delays between the Earth and the Moon or Mars raises a requirement of robust robot performance (for Mars; 22min maximum, 3min minimum). The Camera Space Manipulation methodology provides a computationally less intensive path planning strategy to accomplish semi-autonomous robot operations.; The contribution of the work to the field of vision based robotic maneuvering is the development of a CSM based methodology for flexible linked robots. The work for this dissertation consists of analytical and experimental investigation of the performance of Camera Space Manipulation for a kinematic model of the flexible manipulator which accounts for the gravitation. Trade-offs between camera view parameters and transverse and axial deflection model parameters were investigated. Furthermore for end effector positioning, the conditions for existence of a set of camera view parameters under the presence of modeling errors was explored.