06. Feb 2026
AI Implementation for Predicting Quality Deviations
As part of a project with Zukunftszentrum KI NRW, we investigated how machine learning can help detect quality deviations early and avoid unnecessary rework.
The Challenge:
In production, additional work steps arise when parts need to be "straightened" afterwards to guarantee straightness. This process is time-consuming and costly. The project's goal was to develop an AI model that predicts whether rework is required – already during the initial production stages.
The Approach:
- Analysis of over 4,000 historical production cases
- Consideration of relevant parameters such as material properties, processing data, and process information
- Development of an AI model with a prediction accuracy of up to 90%
The Result:
The AI can predict with high probability when rework is necessary. This allows production times to be shortened and costs to be reduced. The project demonstrates how AI can make an important contribution to economic sustainability.
Outlook:
The developed solution has been implemented as a prototype and can be integrated into live operations in the future. With continuous data input, prediction quality will continue to increase – a crucial step towards digital transformation.
Zurück zur Newsübersicht
