Gesture Concept and Current State’

Zusammenfassung (Abstract) :

The goal of the 3-year EU-project “MOONWALK“ is to evaluate the usability and advantages of an application of robots in astronaut-robot cooperation, which is to be demonstrated in earth-analogue environments, specifically in a sub sea (moon analogy) and desert (mars analogy) scenario. Central element of the planned astronaut-robot cooperation scenarios is the deployment of a gesture based control concept, which potentially allows the user a hands free commanding. As an astronaut wearing his suit may usually being impaired, i.e. not being able to perform movements and actions as being able while not wearing a suit, the gesture control concept would have to focus on minimum effort as a main criteria. As the constraints induced by wearing a suit also limit the number of movements available to the user, the control concept has to effectively make use of a reduced number of available gestures to implement the variety of needed commands. Beyond that the gesture control concept should allow the user to potentially perform additional tasks while issuing gesture commands. To tackle this requirements and restrictions, the control concept presented here features two aspects of special importance regarding the conditions given by the scenario: a posture independent, state based capture of gesture commands and a multilevel input concept. The former allows the user to issue commands independently of performing other actions simultaneously, the latter allows for reusing gesture sequences in different modes depending on context states and formerly issued commands, thus coping with the limited number of movements available to form gesture commands from in the scenario given. First tests with a simplified version of this control concept proved feasibility of the concept and also showed that, using the presented control concept, controlling a complex device like a robot can be performed almost intuitively using gesture commands.