Modelling the Collaboration of a Patient and an Assisting Humanoid Robot During Training Tasks

Abstract

The paper describes an approach for using a humanoid robot as a motivator for patients that recover from stroke. Specific training tasks for arm ability training exist. They were implemented in such a way that the robot gives advices and provides motivational comments. The language DSL-CoTaL was used to specify the collaboration between patient and robot. It allows the visualization and animation of the models with different tools (CoTaSE, CTTE, HAMSTERS). A task pattern was identified for the exercises with a mirror. This pattern reduces extremely the necessary modelling support. It also helps to identify requirements for a domain-specific language for robots, that will allow to generate the necessary code. The language is planned to be an extension of CoTaL.