Published July 2018.
Since only a suitable context enables meaningful analyses, the first task was to automatically assign the data from ongoing production to the data from the outgoing goods inspection. Subsequently, the difficulty was to derive meaningful statements for the production managers and machine builders from the many complex correlations.
The pilot project revealed correlations between different measurement characteristics. This enables the brandgroup to adjust the measurement characteristics even more precisely and to identify quality deviations at an early stage. In addition, the SDSC-BW experts used factor analysis to identify two measurement characteristic groups that provide information about typical defect patterns. With the help of contextual knowledge, the brandgroup can thus significantly reduce the effort and time of future sampling.
Data Innovation Community
Brand Group, Smart Data Solution Center Baden-Württemberg
June 2018 – July 2018