Swiss Researchers Develop System to Transfer Skills Across Robots

A team of researchers at the École Polytechnique Fédérale de Lausanne has developed a system that allows robots with different designs to learn and perform the same task without rewriting code. This system, based on a motor intelligence framework, separates the concept of skill from the specific physical characteristics of a robot. In practical tests, three different commercial robots successfully completed a complex sequence of actions, demonstrating the effectiveness and predictability of this approach.