Through the highly innovative processes introduced by the movement Industry 4.0, application of fully autonomous working processes is ready to be integrated into real-world manufacturing applications. While many tasks still require the human to handle specific parts or to perform certain steps in the production chain, partial assembly may be performed by cooperating mobile manipulators in the near future.
However, many small and medium enterprises (SMEs) fail to adapt to the new trends. Industry 4.0, in spite of its beneficial aspects, is costly to implement, particularly in still low-automated sectors. As part of the SHAREWORK project, RWTH shows that process automation is possible on the low-end scale. Therefore, the Institute of Mechanism Theory, Machine Dynamics and Robotics of RWTH Aachen University has applied the SHAREWORK methodology in a scenario that involves only low-cost mobile manipulators.
RWTH demonstrated how Artificial Intelligence (AI) task planning can be integrated with computer vision systems to achieve a partially autonomous manufacturing process. These processes were applied to a lab environment derived from one of the , further incorporating a mixed team of mobile agents (i.e., human and mobile manipulator).
The targeted SHAREWORK use-case is that of Goizper, a SME located in Spain. As part of the Goizper use-case, screws must be inserted in a rotatory table before they are tightened. The insertion process is a repetitive, low-added value activity that should be automated. The automation solution proposed by RWTH is a two-part system containing only low-budget elements.
First, a vision system uses a 3D camera and an algorithm developed as part of SHAREWORK to automatically identify in which holes screws are already inserted. Based on the outputs of this algorithm, an AI task planning framework is deployed. This framework automatically determines the tasks for the robotic mobile manipulators (e.g., navigate, grasp, place) such that all required screws are inserted. Furthermore, this sensing-planning-acting framework also implements the execution of these tasks in the real-world environment.
Look at how the methods play out in the validation video:
RWTH Aachen University is one of the leading universities for engineering in Germany. The Institute of Mechanism Theory, Machine Dynamics and Robotics, from the Faculty of Mechanical Engineering, participates as a task leader for semantic cognition, also bringing in their expertise in kinematics, sensor vision and task planning.
Stefan-Octavian Bezrucav is pursuing his Ph.D. and is the team leader of the Robotics and Mechatronics team at the Institute of Mechanism Theory, Machine Dynamics and Robotics (IGMR) of RWTH Aachen University. He holds a bachelor’s and master’s degree in Computation Engineering Science. His main research area is the automated task planning for mixed teams of mobile manipulators, humans and robots, spatial reasoning, and complex simulations.
Nils Mandischer is pursuing his Ph.D. at the Institute of Mechanism Theory, Machine Dynamics and Robotics (IGMR) of RWTH Aachen University. He holds a master’s degree in Automation Engineering. His main field of work is computer vision in mobile robotics with special focus on human tracking and navigation for vision confined applications.