Our goal is to understand the principles according to which (semi)autonomous robot systems should be built if they are to act in situations unforeseen by their designers, in environments with uncertainty and change, and be able to understand various forms of novelty. The main objective of RoSy is implementation and evaluation of embodied software components necessary to meet this goal. The most important aspect of the implementation is a multidisciplinary research including the areas of computer science, mathematical sciences, mechanical and electrical engineering. RoSy will thus implement and evaluate integrated and systems oriented demonstrators, using several physical robot platforms with mobility and manipulation capabilities. Our research will be guided by several scientific problems that need to be addressed for realizing two anticipated demonstrators:
The intelligent co-worker
The intelligent co-worker collaborates with humans or other agents; identifying, learning and manipulating known and new objects; capable of identifying and handling errors.
The intelligent site explorer
The intelligent site explorer is context-aware, capable of understanding its environment, detecting and recognizing places, objects and people; working individually or in a multiagent scenario.