Research objectives and content
The objective of the project is the development of an universal architecture for "mobot-brains", i.e., a computational hardware platform for mobile robots. In doing so, the project especially focuses on the needs and concepts in Artificial-Intelligence research in respect to robots. Although the idea of using robots as a basis for AI research is widely spread, simulations and software agents are still dominant. The crucial problem is, that there is no easy-to-handle robotic agent, which can be customized by the researcher. The intended project shall overcome this lack by providing
a) a theoretical framework which helps to design robotic agents on an abstract level and b) a tool-case of easy-to-handle building-blocks to construct a real robot.
This goal will be pursued as follows. Outgoing from an analysis of the existent non-flexible control units designed for fixed applications, common user demands are investigated. Together with the constraints and the possibilities due to the available actuators and sensors this leads to a definition of the requirements to the universal architecture. In addition, the electrical robustness and also the mechanical handling have to be taken into account. Finally, the resulting architecture is realized in hardware. Training content (objective, benefit and expected impact)
As the main benefit of the intended project I will gain deeper insights in the construction of autonomous hardware systems and as a member of the VUB AI lab I will broaden my knowledge in the field of Artificial Intelligence. Links with industry / industrial relevance (22)
The proposed project has an industrial relevance as it can help significantly to boost the increasing usage of autonomous mobile robots in industrial applications, e.g. in sewage system maintenance or surveillance.