Re-Machine: Circular economy for machine parts

At a glance

• Even though steel can already be recycled well, the melting and production processes are associated with considerable CO₂ emissions. The reuse of metallic parts could solve this problem.
• In the Re-Machine project, the scientists are developing a closed parts cycle using the example of bolts in wind turbines and employing self-learning algorithms and a digital twin.
• The end result will be a demonstrator that is easy for users to operate and can deliver clear statements regarding the reusability of individual bolts.
• Among other tasks, Fraunhofer IWES is responsible for the development and implementation of prediction models as well as the construction and evaluation of the prototype.

The challenge

The recycling of steel is a tried-and-tested process: once machine parts have reached the end of their service life, they are melted down and turned into new parts. However, the process is energy-intensive. Even in modern electric arc furnaces, around one ton of CO₂ is produced per ton of steel processed. The reuse of metallic parts could solve this problem. However, there is not currently a market for reused machine parts, as reliable inspection is not yet state of the art. 

The solution

This is where the Re-Machine project comes into play. The aim is to tap the potential of a closed parts cycle using the example of bolts in wind turbines. Bolts are an obvious choice, as they are simple parts that are widely employed and well researched. However, their reuse has not yet been scientifically investigated.

The Fraunhofer-Gesellschaft has already developed a patented process employing self-learning algorithms that is capable of judging the reusability of a part. The scientists will now train these algorithms with a set of bolts produced in the project. This will involve the application of new approaches for the determination of the remaining service life of large bolts (M36 – M72). In addition, new machine learning architectures for data fusion and analysis as well as for classification must be developed that deliver explainable and reliable results for automated decision-making. A digital twin will serve as a central data space.

The end result will be a demonstrator that is easy for users to operate. The bolt will be inserted into the demonstrator, and a binary statement on reusability will be offered within a short time. If the bolt meets the requirements for a new bolt in all relevant properties, then it is assessed as new, regardless of how often it was already installed and how heavily it was loaded. The scientists aim to license the technology to testing companies, machine parts dealers, and maintenance service providers.

Among other duties, Fraunhofer IWES will be responsible for the development and implementation of prediction models as well as the evaluation of the prototype.

The added value

Re-Machine will contribute to saving considerable quantities of energy and greenhouse gases. At the same time, it will help to reduce dependence on raw materials providers in the currently complex geopolitical climate. Last but not least, the expected wave of repowering in the wind energy sector brings with it potential for a large market: in the near future, countless bolted connections will be dismantled, while a growing need for new turbines will lead to an increased demand for bolts at the same time.