Four years of research, four use cases: As part of the “Datenfabrik.NRW” project, industry and research partners jointly developed data-driven solutions for production and logistics. Using technologies such as Motion-Mining®, previously hidden “blackbox” processes were made transparent, enabling AI-powered optimizations in manufacturing and order picking.
How can manufacturing companies remain competitive when faced with labor shortages, supply chain disruptions, and increasing sustainability requirements?
This question marked the starting point of the “Datenfabrik.NRW” research project in 2022, driven by a strong consortium of industry and science. The goal: to develop data-driven solutions that make production, logistics, and enterprise processes smarter, more efficient, and more sustainable.
Nine project partners – including CLAAS, Schmitz Cargobull, NTT Data, Duvenbeck, the Fraunhofer Institutes IEM, IML, IOSB-INA, IAIS, and MotionMiners GmbH – worked together for four years on innovative use cases and technological approaches for the factory of the future.
With a total budget of €14.5 million, €9.2 million of which was funded by the state of North Rhine-Westphalia, Datenfabrik.NRW is one of Germany’s largest initiatives for applied AI.
The project was structured into four key areas, known as “Transformation Areas” (TA). These served as the foundation for developing data-driven solutions and addressed the most critical challenges on the path to the factory of the future.
Each TA was designed to transform traditional manufacturing into a data-based, connected, and AI-enabled production environment. Interdisciplinary teams from research and industry worked on concrete measures, technologies, and use cases that were tested directly in practice.
The four Transformation Areas at a glance:
Within the Transformation Areas “Data-Driven Logistics” and “Data-Driven Manufacturing,” MotionMiners GmbH played a central role in capturing and structuring data in predominantly manual processes. As a technology partner, the company worked with research and application partners to develop solutions for use cases focused on order picking and assembly.
A major challenge in digital process optimization lies in so-called “blackbox” processes – areas where little or no data is available. Manual tasks often lack the structured information essential for implementing digitalization and AI solutions. This is where Motion-Mining® comes in: the technology enables automated and anonymized collection of movement and workload data directly on the shop floor, making previously hidden processes visible.
The data provided the foundation for improved process transparency and served as input for advanced AI models. The technology was deployed across several use cases, including the analysis of logistics workflows and ergonomic stress factors. These insights fed into digital optimization initiatives and supported the development of new tools and strategies.
Motion-Mining® proved to be a cornerstone for data-driven optimization of manual processes throughout the project. In the coming years, these use cases will be expanded and complemented by additional data collection, particularly in assembly, to achieve even greater process transparency.
Over the course of the project, four concrete use cases were implemented, demonstrating the versatility and impact of Motion-Mining®:
CLAAS: Ergonomics and Workforce Planning in Order Picking
Motion-Mining® was used to capture movement and workload data in central order picking, enabling ergonomic assessments and time analysis for different items. These insights formed the basis for a software tool for data-driven workforce planning. Together with NTT Data, strategies for item storage will be developed and implemented based on this data.
Schmitz Cargobull: Digital Process Integration and Plant Layout Planning
Three logistics processes – from storage to order picking to assembly line supply – were integrated into a digital continuous improvement process (KVP). Motion-Mining® enabled detailed mapping of manual workflows, creating a new level of process transparency. The data was used as parameters for a Fraunhofer IML AI tool for staffing strategies and provided valuable input for future plant layout planning.
Ergonomic Analyses: Measuring Stress in Manufacturing
Two additional projects focused on ergonomic stress in manual manufacturing tasks. The goal was to improve working conditions, identify efficiency potential, and optimize task distribution. The findings informed workplace design measures and recommendations for production planning.
Close collaboration between industry and research was key to success. Combining practical experience with scientific methodology enabled solutions that could be directly integrated into daily operations.
Participating companies optimized their processes and provided valuable input for technology and software development. MotionMiners, for example, benefited from real-world requirements that influenced product development – a prime example of productive exchange between application and research.
Another success factor was comprehensive change management: more than 500 employees at CLAAS and Schmitz Cargobull were trained in working with AI.
The collaboration does not end with the project. Follow-up initiatives are already planned with several partners, including expanding data collection in assembly and refining existing AI models. Datenfabrik.NRW has laid the foundation – now it’s time to build on these insights and explore new innovation opportunities.
The project stands as a prime example of successful cooperation between industry and research. It demonstrates how data-driven technologies can make production processes more resilient, efficient, and sustainable – and highlights the importance of strategic partnerships for the future of manufacturing.
Our free factsheet provides you with a comprehensive overview of our hardware and software solution.
Overview of the functions of the MPI Dashboard
Details about anonymization and data security
Technical specifications of the hardware and software
.webp)
).webp)
