Below, we summarize some of the projects we work on and projects which are already finished.
Cooperation between humans and robots has already made promising progresses in the field of stationary robotics - for instance the two-arms robot BAXTER. However, the cooperation in the area of mobile robotics is still at the beginning. Hence, the Professorship of Automation Technology is focused on this research topic together with the Professorships of Media Informatics, Computer Graphics and Visualization, Computer Engineering, and Media Informatics (HTW Dresden).
As scenario we have chosen a construction site - i.e. a GPS-restricted dynamic unstructured environment - on which workers and robots have to cooperate. The most import research topics are mobile robotics, human and gesture recognition, visual, auditive and haptic communication with the robot, visualization of relevant data, and efficient data storage.
The project of “Smart Virtual Worker“ (only in german avaible) funded by the European Union with the European Social Fund (ESF) and by the state of Saxony presents an opportunity to easily replicate established workflow parameters inside a virtual simulation to establish alternative routes, storage, or construction methods during the stage of production planning.
Therefore, we designing a cognitive and emotional model for a digital human worker which simulates a worker’s emotional feelings throughout a typical task inside an industrial setting. In this context, we try to link psychological, biological and artificial inteligence research fundings to a complex model representing a real human worker.
Field of research – Human-Robotic-Interaction: The model aims to enable a robot or any other automated system to perceive, calculate and predict the emotional reaction of humans during a physical stressing work task.
The „StayCentered - an assistance system for air traffic controllers“ (MaCeLot)-project is funded by the Federal Ministry of Education and Research for the next three years. Main focus of the project is to develop a system for team-tasks, especially for air traffic controllers. Therefore, it is necessary to model the interactional behaviors of all real world group members as virtual agents. To get those as realistic as possible, the individual physiological activities of the air traffic controllers are recorded. Sensory inputs (e.g. body posture, gestures, facial expressions, pitch of voice as well as physiological parameters like skin conductance and eye-tracking) allow us to determine whether or not a controller might be stressed or close to a mental overload.
Based on the established emotion model of the "Smart Virtual Worker" project we are going to design an enhanced version including interactional group behaviors and cognitive stress levels.
Field of research – Human-Robotic-Interaction: The model aims to enable a robot or any other automated system to perceive, calculate and predict the mental work load of interacting humans.