In the DEFIne project, the monitoring of progressive dies using artificial intelligence is being researched. As part oft he project, an industry-oriented modular progressive die with integrated sensors is being developed. With the sensor-integrated progressive die, the production process is to be monitored by means of AI-supported models in ordert to obtain information on the condition of semi-finished products, tools and components. From this, strategies for monitoring progressive dies in the industry using artificial intelligence are derived.
Digitization promises enormous growth potential and, especially for small and medium-sized enterprises (SMEs), an increase in overall productivity. Above all, SMEs in forming technology, which are characterized by high productivity and complex multi-level process chains, rarely use the potential of digitized production. Modern production processes can now combine more than ten forming stages on progressive dies, which means that the number of manipulated and disruptive variables and the requirements for process control are constantly increasing. If one also takes into account the data that is widely available in the context of the advancing sensory equipment of production processes, the result is a complexity that no longer allows an analysis by the specialist staff during the process management. As a result, individual product properties are confronted with a large number of setting parameters and process information from the various forming stages of a progessive die, which makes a comprehensive, physically-based process understanding considerably more difficult.
In the running process, it is therefore not possible to make statements about the state of the semi-finished product, tool and component due to the number and complexity of the forming operations. Systems for monitoring stamping and forming processes in the industrial environment are currently based on the definition of limit values and the comparison of process forces with reference states. The use of other process variables (acceleration, tool displacements, component geometry, etc.) as well as the synchronization of such data from different sources is not state of the art.
Digitization and the associated sensor qualification and characteristic value-based modeling based on AI-supported methods have great potential to master these processes, to support specialist personnel in process analysis and process management and thus to increase the overall productivity of metal forming companies.
In order to achieve this goal, a modular progressive tool is developed taking into account industrial standards with integrated sensors and a software platform for networking, structuring and synchronization of company-specific data is being developed and models for describing the state of the semi-finished product, tool and component are built. Here, data-driven analyzes are combined with domain-specific knowledge in order to increase the quality of the AI-supported process models.
From this, strategies and recommendations for action for the model-based monitoring of progressive dies are derived in order to make it easier for companies in the forming technology to get started with AI-supported monitoring of their progressive tools in the future.
The DEFIne project ist funded by the Dist@l funding program. In addition, thanks go to the project partners Intelligent Data Analytics GmbH & Co. KG and Thomas GmbH.