AI-assisted Producibility Analysis
The challenge: The initial setup process of profile wrapping machines is currently done manually by the operator.
The solution: A self-learning AI cycle is established around the digital twin of the machine, which continuously provides recommendations for optimal machine settings.
The added value: Recommendations are provided regarding the manufacturability of new products and parameters for future encapsulation processes.
AI knows the optimal machine setting
Düspohl is considered the most innovative company worldwide when it comes to the development and manufacturing of profile wrapping technologies, surface laminating, and peripheral machines for the wood and plastics industry. At Düspohl Maschinenbau GmbH, the wrapping of wood, metal, or plastic profiles is done fully automatically and intelligently using the “RoboWrap” technology, which employs pressure rollers. On the AI Marketplace, the goal is to optimize the setup process of the manufacturing plant and generate recommendations for the feasibility of products using a digital twin.
Profile wrapping is a process in which a decorative surface is laminated onto a substrate. The wrapping is done using pressure rollers on a profile wrapping machine. Currently, an operator manually positions the pressure rollers when initially setting them up for a profile geometry, learning the optimal configuration. The combination that yields the best wrapping result is then stored and can be recalled at a later time. The robots will automatically reproduce the positions of the pressure rollers based on this stored configuration.
Automated feature extraction and machine learning using AI
As part of the AI Marketplace, experts from Fraunhofer IEM are working together with Düspohl to complete the automation and replace the currently non-automated setup process. Additionally, the feasibility of new product specifications should be automatically evaluated. To achieve this, an algorithm for feature extraction will be developed, enabling the analysis of all profile types at Düspohl based on their characteristics. These features will then be assigned to the individual RoboWrap robots. For each robot, it will be determined in which roll geometry it can best process the assigned feature and where exactly the roll needs to be positioned.
Through a self-learning AI cycle centered around the digital twin of the machine, Düspohl will receive recommendations for optimal machine settings in the future. At the same time, Düspohl’s customers will receive recommendations regarding the feasibility of new products and the parameters for future coating processes. This not only leads to process optimization but also increases efficiency. The insights gained from the project will be generalized for the AI Marketplace to derive an application for the platform.
Let us show you interesting and helpful content in the quickest way possible.