From factory planning to manufacturing to logistics: how can data be used to make individual processes and even a company's entire production more efficient and productive? Pioneering work for this challenge is currently being done by the "Datenfabrik.NRW" project of the leading-edge cluster it's OWL. In the project, the world market leaders CLAAS and Schmitz Cargobull are working together with NTT Data, Duvenbeck Kraftverkehr and MotionMiners as well as the Fraunhofer Institutes IEM, IML, IOSB-INA and IAIS to develop a model for a data-driven factory of the future. The companies are implementing new methods and technologies directly in their production. A total of 50 use cases are to be implemented. This will create state-of-the-art production sites that demonstrate to other companies the added value of data-driven and AI-supported manufacturing. In this way, the project should make an important contribution to securing production in Germany as a business location. At HANNOVER MESSE 2023, interested parties can now obtain comprehensive information about approaches and results at the OWL joint stand.

AI as the key to the data-driven factory of the future

"Artificial intelligence has enormous potential to make machines and plants more efficient and productive. With the data factory, we want to completely rethink production. And consistently collect and utilise data from factory planning to production and logistics. Manufacturing companies throughout OWL and beyond can benefit from the results and experience," says Günter Korder, Managing Director of it's OWL Cluster Management. The "Datenfabrik.NRW" is by far the largest project of the leading-edge cluster it ́s OWL - Intelligent Technical Systems OstWestfalenLippe. A total of around 200 companies and research institutions are developing new technologies for machine learning, platforms and the working world of the future.

For example, for Claas, the agricultural machinery manufacturer involved in the project, the Datenfabrik.NRW is an important building block for the conversion of combine harvester production. "The Datenfabrik.NRW project has come at exactly the right time and is an excellent fit for our plant modernisation," says Stefan Schulte, Managing Director of Claas Selbstfahrende Erntemaschinen GmbH.

The world's most modern combine harvester production is to become even more intelligent

CLAAS has invested 44 million euros in the creation of a completely new 15,000 square metre hall for the world's most modern combine harvester production at the heart of its Harsewinkel headquarters. In the "Datenfabrik.NRW" project, digital solutions are now to be developed that will ensure more quality, efficiency and better jobs - such as AI-supported production planning.

Mastering increased complexity

Individual customer wishes, labour shortages and the desire for the most energy-efficient production possible are just a few target variables that need to be taken into account in production planning. The often implicit knowledge of a production planner and information management by means of Excel lists quickly reach their limits. "Companies are increasingly reaching a point where they are dependent on data consistency and artificial intelligence in production planning in order to be able to master the increased complexity," says Dr Arno Kühn, Head of Strategy, Research and Development at the it's OWL technology network.

In search of suitable algorithms

In the project, researchers from the Fraunhofer institutes are working with companies to find out how algorithms can support planning by analysing data faster and better than humans. In the process, the algorithm searches for the best planning steps to meet goals that the company has previously set - from minimising bottlenecks to reducing costs to low energy consumption.

Scaling AI solutions quickly

A total of 50 use cases are to be implemented and harmonised with each other. In the process, the new technologies and methods will be tested directly in production so that the effects on processes, IT and organisation can be seen immediately and adjustments can be made if necessary. This approach offers great added value for both the companies and the research institutions. "We want to avoid isolated solutions and ensure that the developed AI solutions scale quickly and can be adapted to other use cases and locations. The data factory creates structures and processes in the company that make it possible to recognise the potential of AI and anchor it sustainably in the company," explains Kühn. The knowledge gained about the barriers and solution approaches in the introduction of AI solutions will be transferred by the research institutions to future projects in the context of industrial transformation.

Data factory expertise: factory planning goes digital

AI-supported production is also a topic at the leading manufacturer of semi-trailers, bodies and trailers in Europe: Schmitz Cargobull. The company is consistently pushing digitalisation both in production and in its product and service range, especially as the industry is highly dependent on economic fluctuations, which requires rapid adaptation to rapidly changing conditions, especially in production. Schmitz Cargobull's main production sites are in the Münsterland region - with the plants in Vreden and in Altenberge. With the "Datenfabrik.NRW", the plant for refrigerated box production in Vreden takes a pioneering position with regard to the digitalisation of the production network.

"We will test the findings that we gain through our participation in the project for production, production planning and production logistics directly at the Vreden plant location. The Vreden plant will then take on a model function in our international production network for the further development of production as a whole," says CEO Andreas Schmitz.

The company is planning a plant expansion for the Vreden site. The planning is taking place digitally with the help of the "Datenfabrik.NRW". "We are implementing the planning of buildings, production areas, workplaces and plants digitally to the greatest possible extent even before realisation," says Stefan Cramer, head of the Vreden plant. This would not be possible without support. "So far, we haven't done that and are not yet technically set up accordingly," says Cramer.

Benefiting from exchange

With the support of the experts from the "Datenfabrik.NRW", however, the company can tackle the challenge. "We get access to knowledge and to partners such as the Fraunhofer Institutes, and of course we also benefit from the exchange with other companies such as CLAAS, which have already faced similar challenges," says Cramer.

3D simulations facilitate planning

The advantage of digital planning of the factory expansion: with a 3D model, it is possible to simulate how the factory will function when it is up and running even before it is built. This allows bottlenecks in production and material flows to be identified in good time and optimised accordingly. Artificial intelligence can support the creation of virtual factory models based on a real factory. A 3D reconstruction algorithm can process various images or videos of the real factory floor and merge them into a virtual model. In addition, object recognition algorithms can be used to automatically identify machines and work areas and process them for a 3D model. With this information, companies can improve their processes on the factory floor, thereby increasing their efficiency and productivity.

Transfer to small and medium-sized enterprises

Other companies should also be able to use the results and experiences from the project to make their production fit for the future. To this end, the technology network it's OWL and the state initiative KI.NRW will implement extensive activities. For example, with tours of the new factories and workshops with the project partners, transfer projects and qualification offers. At HANNOVER MESSE 2023, interested parties can find out about areas of application and their effects at a model factory at the OWL joint stand.

Video