Search

Reliability, flexibility and minimum processing time are the key production requirements for sustainably competitive manufacturing. A high degree of flexibility often requires adjustments to both the individual robot-operated systems and the mobile robots used on the process side, resulting in an automatic increase in setup times. A uniform manufacturerindependent programming method for robots would radically reduce programming time. This work describes the development and implementation of a programming concept for mobile robots which allows the use of uniform commands independent of the manufacturer. The resulting programming library breaks these commands down to a level where autonomous code can be generated for the individually used robot, eliminating the need for the user to deal with complex programming tools. [ABSTRACT FROM AUTHOR]

Pick and place tasks are one of the most typical applications in robotics. Conventional robots perform a fixed task which restricts the workspace in terms of flexibility. With development in mobile robotics [1] and automation it is necessary to introduce sensor systems that enable a more flexible approach. This paper demonstrates a computer vision system which uses a colour-based approach to object detection and estimation of its 3D world coordinates based on a 2D image. The system is used to enable a robot arm to pick up an object and place it on a CNC-mill. Based on the use case the vision system can tell the robot if an object has been detected and where it is relatively to the camera. Using the robot operating system (ROS) and OpenCv, image thresholding is performed for detecting the object. The Canny edge detection algorithm is deployed to mark the object with a bounding box. The retrieved information from those operations is used for the 3D pose estimation. Using a linear calibration, the error of the system can be reduced, which results in an accuracy of 0.09cm. [ABSTRACT FROM AUTHOR]

In this paper, we investigate reinforcement learning model explainability through a pick and place task. Two robots with three degrees of freedom learned to solve the pick and place task in simulation as well as reality. To investigate the explanatory factors implicitly learned by the models, we derive robot parameters, i.e., the length of the robot segments. To overcome the black box nature of reinforcement learning models and provide a physical explanation of the results, the robot dimensions are derived from the learned reinforcement learning model and compared to the real dimensions. The hypothesis in the presented work is that converged reinforcement learning models must learn the robot parameters implicitly in order to learn a task. This transforms black box models into white box models, where each model’s decisions can be interpreted. Our experiments show that robot parameters can be derived from learned models and that the chosen reinforcement learning model implicitly learns physical context. In order to create robust and trustworthy AI systems for intelligent factories, we suggest that a physical interpretation of all black box models must be done. [ABSTRACT FROM AUTHOR]