Intelligent production Researchers achieve breakthroughs on the path to Industry 5.0

Related Vendors

Technische Universität Ilmenau

FAULHABER Drive Systems

Taiwan Machine Tool & Accessory Builders' Association (TMBA)

Wuxi InfiMotion Technology Co., Ltd.

In the manufacturing and assembly process of the future, humans and robots should be able to work hand in hand. However, the machines will need to become significantly smarter than they currently are.

The Technical University of Ilmenau, Germany,  has successfully completed its research project "Engineering for Smart Manufacturing (E4SM)". According to a statement, the researchers were able to achieve several breakthroughs on the way to Industry 5.0 within the framework of the project. With Industry 5.0, the human being is to be placed at the center of manufacturing processes in industry and crafts. Humans and machines will be aligned with each other with the help of artificial intelligence, so that new manufacturing tasks can be solved more quickly and efficiently than before.

At the start of the project five years ago, the researchers set out to develop AI-supported robots that autonomously coordinate their actions with humans. The requirements and particularities of small and medium-sized enterprises' manufacturing and assembly processes were the focus. According to the researchers, four milestones were achieved during the project.

3D multi-view stereo system for safe collaboration between humans and robots

In order to be able to ensure safety when working with a robot, the robot must be able to capture the actions of humans in the assembly process in three dimensions. For this purpose, the researchers used multimodal imaging, in which thermal images were combined with RGB color images and with 3D point clouds to form highly accurate large-scale three-dimensional representations of the spatial environment. In the innovative sensor-based 3D multi-view stereo system, the various cameras calibrate themselves robot-assisted.

Recognition of assembly action and corresponding autonomous assistance

Without having to undergo lengthy training, the robot should monitor the progress of assembly and independently carry out recurring tasks, it continues. The researchers developed a robot that uses AI to recognize the actions of human employees and the tools they use. This allows it to track the progress of assembly and decide autonomously where it can help. The robot can also locate and grasp any objects in the operating environment that it was previously unfamiliar with. According to the current state of the art, this only works for previously defined, known objects.

Flexible and cost-saving AI-supported laser beam welding

Up until now, intricate and expensive clamping devices had to be used in laser beam welding to fix the sheets that were supposed to be joined together. With AI, the welding process can now be analyzed contactlessly. The system can predict when a gap will occur. By controlled compression of the sheets using robots, this can be avoided, so that the laser continues to hit both sheets and can thus weld them together. The AI predicts the force that must be expended to optimally adjust the position of the sheets for the welding process.

Interactive visual tool chain for small and medium-sized enterprises

The researchers also developed a visual tool chain. This is designed to assist small and medium-sized enterprises in developing efficient and safe applications that can be used in their own production. As this doesn't require any detailed technical knowledge, the development process is significantly simplified for users, according to the researchers.