A resumé for products: Increased efficiency and added value through closed life-data cycles

Until now, it has been impossible to systematically record product data from the raw material stage through to recycling: the value chain is too complex for this and the data flow is not continuous enough. However, this leaves significant untapped potential for optimizing production. As Fraunhofer’s DMD4Future project is now showing, eResourcing can provide solutions that address this by collecting and connecting data at different points in the process chain. What makes this remarkable is that the data can be used to optimize many different stages of the value chain.

Digital connectivity in industry has its charms: cameras that automatically monitor the quality of components on assembly lines; sensors that independently call the maintenance technician when a tool is worn out; production systems that automatically order supplies when components run low. These days, more and more devices are being equipped with powerful microprocessors and an Internet connection so they can communicate with one another. This trend toward the complete connectivity of machines is well known as Industry 4.0. But despite all this progress, one thing is still almost impossible: fully capturing data about a component from the raw material stage through production and use to the end of its life. “To date, no industry has been able to obtain and use information systematically over a product’s entire life cycle,” says Dr. Stefan Gerth, head of department at Fraunhofer IIS. “That’s unfortunate because it means we’re missing out on huge potential for optimizing production, improving product quality, increasing sustainability, and ultimately saving money.”

Working together with experts from 15 Fraunhofer Institutes on the Fraunhofer joint project entitled DMD4Future – short for for Digitalized Material and Data Value Chains for the Future – Gerth has now investigated how to capture and use the wealth of data arising along the value chain until the end of a product’s life. To this end, the team closely analyzed topics including the production of cast parts for vehicles and engines.

Today, before cast parts are delivered, they are inspected in the foundry using X-rays, ultrasound and other methods. This includes examining them for pores, cracks or flaws that could later lead to damage. The results of these tests are summarized in a test report, which is then often delivered to the customer in the form of a PDF file together with the cast part. This means all the detailed information that was previously obtained by X-ray or ultrasound doesn’t accompany the part on the rest of its life’s journey. However, this data might be tremendously valuable: should the component fail at some point, an analysis of the X-ray data collected could reveal which weak point may have led to the failure. “Over time, this could help manufacturers work out which microstructures are more likely to fail,” Gerth explains. “These findings could then be fed back into the foundry process to improve it.” Manufacturers might conceivably also choose not to use structurally identical cast parts with small flaws in high-performance engines, but rather in engines that are exposed to less stress – such as those that operate at lower temperatures. Gerth says, “These examples show that there are many ways to use digital information about a component or product, if only it were captured.”

Using data multiple times

This kind of multiple use of data is part of the eResourcing strategy developed within Fraunhofer’s DMD4Future project. In this case, the idea is not to let data generated in production lie forgotten, but to market it. The challenge is to transfer information that originates at a certain point in the value chain to subsequent stages of production. “In a sense, we have to link several data clouds together for this,” Gerth says. Analysis of the various steps in the foundry process has shown that this can work; in particular it calls for expertise on interfaces. Among the things Gerth and his colleagues rely on is the “OPC UA” international interface standard, which was developed for Industry 4.0.

One specific solution for an end-to-end data flow across a product’s entire life cycle is to give the product a detailed digital twin that incorporates data from all stages. This can include sensor data that shows how much stress a component is subjected to during operation. In the future, combining this information with the knowledge gained from testing as part of the foundry process would make it possible to estimate whether a component could continue to be used beyond its intended service life – if, say, it had been subjected only to low stresses during operation.

Feeding real information back into production

“The beauty of this eResourcing concept is that we’re bringing together real data from the process to generate a variety of solutions that we feed back into the real world,” Gerth says. Such an all-encompassing data set would also be useful at the end of life, when a product is to be recycled. Since a cast component’s digital twin would also contain data on the quantities of raw materials used, the part could be recycled in a much more targeted manner – rather than simply being thrown on the scrap heap.

Even today, foundry processes are still characterized by how much manual work and empirical knowledge are required. Foundry tools such as molds wear out over time due to the high temperatures. The equipment must be readjusted accordingly to maintain the desired quality. Experienced technicians have the relevant knowledge – but if someone drops out due to illness or retirement, this knowledge is lost. “eResourcing now offers the opportunity to record machine parameters in much greater detail than before and to link these with information about product quality,” Gerth says. “That would be of help in controlling the plant.”

The DMD4Future project’s analysis of the foundry process has shown that such full digitalization is feasible. In the future, the principle is set to be extended to other production processes.