CNH improves quality control of combine harvesters with data-driven end-of-line testing

In the production of combine harvesters, thorough quality control at the end of the production line is essential. For CNH, this raised a clear challenge: how can deviations in rotating components — such as bearings, imbalance, or chain-related faults — be detected more quickly and more reliably in newly produced machines? Traditional end-of-line tests provide a general view of how a machine performs, but they do not always make it easy to identify these specific issues early and in a systematic way.

This is where the collaboration with Flanders Make added value. Flanders Make identified which sensors and which features in the measurement data were best suited to monitor the condition of new CNH combine harvesters during the DRI test (dynamic ride-in or end-of-line testing). The focus was on detecting faults related to bearings, imbalance, and chain issues in different components of the machine. These features were then integrated into a Power BI dashboard to make the condition of individual machines visible during the end-of-line test.

Improved end-of-line tests through data

The approach starts from measurement data collected during the end-of-line test using carefully selected sensors. These sensors are temporarily installed at specific positions on the combine harvester to capture relevant data and are removed again after the test. Based on this data, features were developed that make it possible to identify and track potential deviations. This gives CNH a much clearer view of the condition of each individual machine at the moment it leaves the production line.

The sensor data collected during testing is sent to a cloud environment, where it is further processed so that it quickly becomes visible whether a machine meets the expected quality standards or whether targeted action is required.

Data creates double value

This data-driven approach creates value in two ways. On the one hand, it enables more thorough quality control of each individual machine at the end of the production process. On the other hand, the systematic availability of data across many produced machines also provides valuable input for design and R&D. By analysing trends and recurring deviations more effectively, CNH can not only detect faults faster, but also implement structural improvements in design and production.

For CNH, this represents an important step towards smarter quality control at its plant in Zedelgem. By combining sensors, data analysis, cloud processing, and clear dashboards, end-of-line testing becomes not just a control step, but also a source of insights to further improve machines and processes. This case shows how research knowledge, digital infrastructure, and industrial implementation come together in a concrete application with direct impact on quality and continuous improvement.