Researchers with the European research project XPERO developed a machine learning method, which enables robots to learn fundamental mathematical and physical concepts by experimentation and observation.
The robot’s sensor data is recorded while it moves and creates a model, which allows the robot to predict how the objects in its vicinity will change their position in relativity to the robot. Although learning a coordinate system might look like a rather basic research program it represents a significant technological relevance.
This XPERO project,coordinated by the Bonn-Rhein-Sieg Univ. in Sankt Augustin, Germany, lays the foundation for the technology to build the next generation of so-called service robots, which could clean houses, mow lawns or do your household chores.
XPERO’s learning robot was presented in April 2009 during the FET’09 conference (Future and Emerging Technologies) in Prague, Czech Republic.
© Flavia Westerwelle