Given the limited amount of raw material combined with today¿s throwaway society, the reuse of goods is from crucial importance. Hence, the disassembly of products is an essential part of a product¿s life cycle. Disassembly technologies currently in the focus of research represent a basis only and cannot answer well to the need for autonomous disassembly processes. Concerning the production processes sufficient knowledge about the process itself and the involved machines as well as product parts is available due to parts lists and documented assembly steps, but regarding disassembly processes this knowledge is missing. Therefore, to be disassembled efficiently the corresponding component assemblies have to be perceived and determined by comparison with a knowledge base. Optical tracking in combination with knowledge bases and the ability of mobile robots to autonomously learn and act promises to be a solution to the problem of autonomous distributed disassembly. New technologies like Multi-Agent and semantic systems lead to a decentralized concept with artificial intelligence and promises robustness against disturbances. The usage of semantics copes with the challenges of communication between heterogeneous systems. The aims of the project are to develop a concept for distributed intelligent disassembly of component assemblies with mobile robots and to encourage pupils for the research in the field of autonomous robotics. By participating in the Botball-program the young scientists should be able to recognize and learn easily about the problems of robotics in disassembly processes. This program is the optimal basis for further cooperation with the possibility of annual tournaments at the Vienna University of Technology.