GESTIK - Neuronale Architekturen zu GESTikbasierten Interaktion und Kommunikation mit dem mobilen Robotersystem MILVA


 

The control of autonomously acting, mobile robot systems is an internationally and increasingly also nationally strongly promoted research direction with broad fields of application in industrial automation and in the service sector.

In the service sector in particular, a new generation of systems has recently been developed under the term "service robots", which are intended to provide everyday services and assistance in direct cooperation and interaction with humans.
However, extremely high demands are placed on the performance of the control and sensor systems required for this as well as on the intelligence and flexibility of the "user-robot interfaces", especially with regard to the ability of communication and interaction between user and service robot and thus acceptance by the user.

Within the framework of this research project, adaptive neuronal architectures for non-verbal, visually based communication between user and robot are to be developed, with the focus on action-relevant gesture interpretation and visually guided local navigation of the robot to the user.

The mobile robot system MILVA with on-board neurocomputer and multisensory data acquisition, which is financed from state funds, is to be used as the research platform for the project.