Innovative, camera-based assistance systems

In the future, users will need more and better assistance systems, because production in the factory of the future will become more complex, the workforce will grow older on average, and new products and processes will be introduced continuously. Assistance systems prepare information for the user, evaluate alternative courses of action or support the execution of decisions. Used in a goal-oriented manner, they improve the flexibility of people and technology, thus enabling decisive competitive advantages to be achieved. To achieve this, it is necessary to abstract from concrete technical core competencies and to view industrial business processes as a whole. We pursue a consistent concept from the recording of machine data in real time, through their processing and generation of process knowledge, to the visualization of user-specific information and instructions for action.

We develop assistance systems that our customers can use flexibly in production because they are modular. This includes technologies that support users on-the-job, even at short notice for less predictable work activities, e.g. in the form of augmented reality or virtual reality assistance systems. When we develop such user-friendly systems, we follow an iterative design approach: we accompany the development of hardware and software by continuously testing the prototypes in qualitative and quantitative studies close to application. Using machine learning methods, we develop systems that can also be adapted to the respective usage context and better integrate the user into the production process.

With the help of self-developed camera- and video-based procedures, we give technical systems the ability to perceive the user and thus support him more effectively. In detail, we have the following technology modules at our disposal:

.

• Our camera-based facial analysis provides information on the identity, age and gender of the user in order to personalize and automatically adapt user interfaces to individual needs.
• With the help of our gaze and head rotation measurement, an IT application 'understands' what the user's attention is focused on; for example, it can establish the reference to the target object for which input is to be made or assistance offered.
• Our methods for measuring posture provide information about movement, physical activity, ergonomics and the strain on the user.
• In addition, the associated hand and arm recognition serves to implement intuitive interaction solutions, such as intuitive hand poses or pointing gestures.

 

A goal is to combine these components with small implementation expenditures flexibly to application-specific assistance systems.