Growing challenges in quality assurance
Perfect product quality is a trademark of German medium-sized businesses. This is often associated with considerable effort in quality assurance and high costs for production rejects. Simultaneously the challenges in the analysis are constantly increasing: innovative products at the limits of what is technically feasible, a high number of variants, small batch sizes and ever shorter product life cycles. The classic manual “analysis of the defective individual part” by experienced engineers often reaches its limits: when acute problems have to be solved quickly, or when the causes lie in a combination of effects. The human expert can sometimes discover a fault in the process “at first glance”. But when defective parts pile up on the lab bench and things have to be done quickly, a complementary approach is needed.
With data and AI to a complementary approach
Here, a holistic view of data, combined with the right AI, can quickly deliver the right ideas. As a basis for such AI assistance systems, you need a consistent flow of production data representing as many influencing factors on quality as possible. Modern Manufacturing Execution Systems (MES), supplemented by sensor technology (IoT) and ERP data, can usually achieve this. The reward for the effort involved is not only improved diagnostic capabilities, but also, through intensive use, higher data quality overall. Above all, however, such solution projects are also catalysts for completely new collaboration models in which experts find new solutions based on data across processes and departments. Whether in discrete mass production or in the process industry, data-driven quality assurance with AI support can thus make an important contribution to shorter start-up phases for new products and to faster diagnosis of acute error patterns, thus sustainably reducing quality costs and increasing the contribution margin.
The resulting advantages are:
- More transparency in production
- Shorter cycle times in quality assurance
- Increased diagnostic capability and more collaboration throughout the production
- Higher data quality through intensive use
- Faster and permanently less waste