Fraunhofer Institute of Optronics, System Technologies and Image Exploitation IOSB

The Fraunhofer Institute of Optronics, System Technologies and Image Exploitation IOSB is Europe’s largest research institute in the field of image acquisition, processing and analysis. In addition, the Fraunhofer IOSB conducts application-oriented research in the fields of multisensory systems, bulk material sorting as well as control and monitoring of industrial production processes, among others. For several decades, robotics has been one of the main topics. Here the focus is on algorithms for localization, mapping and motion planning for mobile robots in unstructured terrain as well as on mobile manipulation and safe human-robot interaction. The Fraunhofer IOSB is headed by Prof. Dr.-Ing. Jürgen Beyerer.

Autonomous loading process of an autonomous transport platform © Fraunhofer IOSB

Coordination of ROBDEKON

The ROBDEKON Competence Center is coordinated by the Fraunhofer IOSB. In addition to organizational tasks for project-internal cooperation, an important task is to provide the Competence Center with an infrastructure that enables third parties to easily establish contact and participate in the ROBDEKON network. A coordination office was therefore set up as a central contact point.

Research focus in ROBDEKON

Within the framework of ROBDEKON, the Fraunhofer IOSB focuses on research into autonomous heavy construction machinery. In addition to the automation of such systems, the development of innovative autonomy capabilities is a major focus. This includes, e. g., algorithms for multi-sensory perception of the surroundings, for precise localization of the systems, for planning an optimal motion strategy, and for performing the respective useful function. As an initial application scenario, the autonomous removal of contaminated earth layers is considered.

The heavy construction machines were taught several necessary skills to establish autonomy. Machine learning methods were used to implement environment sensing, ego-motion estimation, and the controlled execution of an earth excavation process. Further, the platform is capable of planning and executing driving maneuvers while reacting to obstacles. Obstacle avoidance for the manipulator arm has also been implemented. The ability to perform an earth excavation, which is rather simple for human operators, is the basis on which further autonomy capabilities are added.