As part of the large-scale research project X-Forge (Everything as a Service) funded by the state of Baden-Württemberg, tool manufacturer MAPAL is the consortium leader of the sub-area Productivity as a Service (ProdaaS). MAPAL collaborates with machine builder F. Zimmermann and measurement technology specialist Blum to provide customers with comprehensive solutions. Another project partner, Fraunhofer IPA, has taken on the task of mediating the underlying business model between the partners and determining the customer value assessment. A first pilot project has been completed after a total duration of three years. Based on this, the digital service offering is to be gradually expanded in the coming years.
The members of the working consortium already monitor processes through many sensors that provide corresponding data. However, existing systems only offer a limited view of the overall system consisting of machine, tool, and workpiece. Although highly complex expert systems are available on the machine, these individual components are not integrated with each other. Bringing them together is practically impossible. Consequently, it is quite difficult to analyze problems that have occurred afterward or to optimize ongoing processes.
Structured knowledge base for process planning
A CAM process planning with tool selection, path planning, and choice of process parameters is a key lever for increasing productivity. However, when machining a component, further adjustments are usually required on the machine to achieve optimal results. Today, the planning area is largely decoupled from machining. Insights gained on the machine do not automatically flow back to process planning. The inadequate feedback hinders learning effects and ensures that the machine operator starts from scratch again with the next planned drilling process. CAM process planning simply lacks a structured knowledge base from practical application.
The project partners contribute different information to the 'Productivity-as-a-Service' offering for stable and efficient machining processes. The wear condition assessment of the tools is relevant to understand the condition in which, for example, a bore was created. It also provides information about the current tool life and allows for tool life predictions.
Blum extracts the relevant data into a dedicated software service, making it usable. These measurement data are compared with a knowledge database from MAPAL to check in a process assessment whether the parameters set on the machine align with the manufacturer's specifications. In parallel, the condition assessment of the machine from F. Zimmermann is included, providing information about the spindle condition, for example. A higher-level software component ultimately correlates the various sources and provides information in a structured manner. In the case of fault condition assessments, the user can determine the causes at the push of a button.
Pilot project Tool Performance Optimizer
In the pilot project that took place at Karl Walter Formenbau, the focus was on a Tool Performance Optimizer. This allowed the user to correct the setting of process parameters for drilling in case of deviations, reducing downtime due to breakage or unplanned tool changes. Furthermore, the database offers the possibility to build on structured, methodically collected, and evaluated experiences from the past, even in the planning phase for new, previously unknown applications using similarity search. The Tool Performance Optimizer is to be marketed through two sales channels. In addition to the classic solution business with a one-stop offering, it will also be available on a large platform that is based on Gaia-X as a data ecosystem.
The goal of the next development step is to capture deviations in an assistance system and warn the user during operation with a traffic light system. For the coming years, self-learning services for autonomous parameter optimization and intelligent CAM process planning are also planned. Ultimately, by 2029, a continuous automation from drawing to finished component should also be possible.
Contact:
