robominds Blog

News from the world of AI robotics
Jul 2023
Tech News
Photocredits: robominds

Teachable robots for the production in Bavaria: robominds becomes part of the project "KoKIRo"

Bavarian know-how combined: The starting signal for the KoKIRo project has been given! An association of regional companies (including robominds) and research institutes starts its joint project in which an adaptive robot assistant for production is to be developed. Bavarian know-how is thus combined at a decisive point in the megatrend of robotics, promoted by the Bavarian state government. The Bavarian Minister of Economic Affairs, Hubert Aiwanger, visited Kaufbeuren in person to hand over the official notification of funding for a pioneering Bavarian research project.

The next step in the evolution of assembly robots

The project partners are pro-micron GmbH, the Munich-based AI robotics company robominds GmbH, the Fraunhofer Institute (IGCV, Department of Digitalization and AI in Production), and the University of Augsburg (Chair of Mechanical Engineering). The goal is to develop an Industry 4.0-capable, cognitive and flexibly configurable robot-assembly-system in close exchange in Kaufbeuren in the building of pro-micron GmbH. For this purpose, the project is not only to make fundamental development contributions for the KoKIRo, but also to demonstrate the benefits and performance of the KoKIRo in so-called use cases at the end. KoKIRo is to recognize the assembly task itself, (re)configure itself automatically and control and execute the assembly process sensitively. Instead of having a robot as a partner in production, which cannot make its own decisions and is merely an extended arm, the KoKIRo will configure itself independently and learn the process, e.g. how much pressure or which tool it should use. Here, robominds will play a central role: with robobrain® (the brain for industrial robots) and smart AI skills, KoKiRo will be given the needed abilities to perform assembly processes that are finely tuned and controlled in real time.

The story with the screw

When a person screws a screw into a thread, he or she intuitively notices when it becomes jammed. He can adapt individually to each new screwing process - sometimes he must screw tighter, sometimes with more feeling. A robot cannot do this. A robot can be specifically trained to carry out a certain screwdriving process in the best possible way. But when it comes to a new screw, for example, it must be retrained and readjusted. In addition, all holes are the same for the robot, but in reality, there are tolerances and deviations. Thanks to the combination of artificial intelligence and sensor technology (feeling), the KoKIRo should be able to retrain itself or behave according to the actual conditions. If the screw threatens to tilt, the KoKiRo should notice this and correct it "with feeling”.

About the project partners

Prof. Dr. Markus Sause, Professor of Mechanical Engineering, is involved on behalf of the University of Augsburg. The University of Augsburg is working on the interfaces between sensing, moving and AI control, with the aim of providing and reliably using the data in real time, of the right quality, securely and in the right place. For Fraunhofer IGCV (Department of Digitization and AI in Production), Prof. Dr.-Ing. Johannes Schilp is part of the project. IGCV is creating the scientific basis for setting up the system architecture, the components and their information exchange in such a way that KoKIRo becomes an universally usable 4.0 system. With robobrain® robominds is giving KoKIRo the needed intelligence and abilities. Within the project, Pro-micron GmbH will research the extent to which it is possible to design standardized mechanical and electronic functional elements as well as equip them with decentralized intelligence on site, that can fulfill the requirements for sensitive real-time control. In addition, the company azero GmbH, a proven expert in application-specific adaptation and integration of the technology to be researched, will be part of this project.

Maurice Brodhun
Head of Marketing