| dc.description.abstract | The progress of computer science and mechanical and electrical engineering in the field of robotics has increased the applicability of robots for replacing human resources in the performance of repetitive and dangerous tasks. A lot of today's work on robots for oil platforms, manufacturing plants and other facilities providing potentially hazardous environments is based on adjusting the traditional industrial robots to operate in fixed coordinate systems. There is a need for enabling robots to move freely on walls and similar structures. One of the major challenges of wall-climbing robots is attaining a secure and reliable grip to the wall. The object of this master's thesis has been to determine the feasibility and applicability of a wall-climbing robot that adheres to the wall by gripping on to bolts, with a control system enabling an operator to easily lead the robot over a considerable distance. In order to do so, KlimBot, a wall-climbing prototype for bolted walls, has been designed, built, programmed and tested. KlimBot has been able to successfully climb horizontally and vertically on designated test walls. Additionally, a control system was implemented; enabling path generation and traversal over several bolts. The test results achieved by KlimBot suggests that there is a great potential for achieving a highly reliable adhesion with the proposed approach. However, if such a wall-climbing robot was to be commercialized, the assumptions made in order for KlimBot to achieve satisfactory functionality would not hold. After discussing options for optimizing KlimBot's design and implementation, the thesis is concluded with a discussion concerning the tough challenges that have to be conquered in order for commercialization to be feasible. | eng |