Today's industrial robot systems often operate in environments designed specifically for them. The goal of research at the Research Institute of Cognition and Robotics (CoR-Lab) and the Center for Cognitive Interaction Technology (CITEC) at Bielefeld University is to further develop cognitive robotic systems in such a way that they can easily adapt to changing tasks, different environments and people with their individual characteristics.
The first example shows a cognitive robotics system for highly flexible industrial production in an assembly scenario. The potential of model-driven software and system development for cognitive robotics is demonstrated by means of an automated terminal block assembly in switch cabinet construction. The cell operator is to be enabled to specify different assembly tasks using reusable and combinable task blocks with which the task is ultimately realized. The methods shown are easily transferable to other applications of robot-assisted assembly, loading and unloading of machines.
The second example focuses on the use of cognitive robotics for intralogistics. Tools and production parts often cover long distances in industrial production plants. As soon as a workpiece is defective or a process has to be adapted, assembly lines are often too rigid. With a flexible means of transport such as a mobile robot and corresponding perception and planning components, production chains can be made dynamic. The AMiRo robot developed at CoR-Lab serves as an experimental test vehicle for questions in the field of intralogistics and mobile robotics. These mini-robots with a high functional density allow the resource-efficient use of inference algorithms that have been trained using machine learning methods. The special feature: With the help of their sensors and the learned behaviours, they interact independently with their environment and work particularly flexibly. The methods and algorithms for environment perception and behavior control can also be ported to larger platforms or AGV systems.