A data driven method for automated detection of anomalous work pieces during a production process

Abstract

The method and system for detection of anomalous work pieces during a production process, comprising computing for each production process step of the production process at least one deviation data signal for a target data signal of a target work piece with respect to reference data signals recorded for a corresponding production process step of a set of reference work pieces, wherein the deviation data signal comprises a number of deviation data samples for different production time steps, t, or path length steps, 1, of the respective production process step; performing a stepwise anomaly detection by data processing of the at least one computed deviation data signal and a process type indicator indicating a type of the production process step using a trained anomaly detection data modelto calculate for each time step, t, or path length step, 1, of the production process step an anomaly probability, p, that the respective time step, t, or path length step, 1, is anomalous; and classifying (S3) the target work piece and / or the production process step as being anomalous or not anomalous on the basis of the calculated anomaly probabilities, p, of the time steps, t, or path length steps, 1, of the production process step, wherein if the target work piece and / or the production process step is classified as anomalous, a warning signal for a user and / or a control signal for a production machine can be generated automatically.

Description

technical field [0001] The present invention relates to a data-driven method for automatically detecting abnormal workpieces during a production process, and in particular to a data-driven method for automatically detecting abnormal workpieces by a quality control system of a production facility. Background technique [0002] Monitoring the quality of running production in a production facility is essential to delivering manufactured products and ensuring consistent quality of manufactured products. During high-volume production, it is nearly impossible to implement extensive quality control on every manufactured product delivered to customers. In most cases, a two-level quality control is implemented. In the first level of quality control, superficial quality control is usually performed by the machine operator, resulting in high coverage but lacking in precision. In the second level of quality control, predetermined in-depth quality control by full-time personnel can be ...

AI Technical Summary

Problems solved by technology

Third, conventional existing quality control systems rely on the assumption that non-obvious defects on workpieces occur serially rather than exclusively

Fourth, conventional methods of quality control are entirely human-driven processes in which the quality of control varies naturally with an individual's human experience and enthusiasm, as well as his or her mood of the day

Images