Supervised Anomaly Detection Model Collection
Machine Learning and Deep Learning models to detect anomalies using time series data.
Supervised anomaly detection using ground truth quality data and manufacturing sensor data from extrusion and blow molding procedures.
Features have been weighted according to the expert knowledge provided by the pilot.
These models have been implemented in the context of the knowlEdge project and demonstrated for Kautex, a multi-industry company leader in designing and manufacturing plastic fuel systems for automobiles and light trucks, including blow molded solutions for conventional plastic fuel tanks and pressurized plastic fuel tanks for hybrid vehicle applications.
Anomaly detection is the process of identifying patterns or instances that deviate significantly from the norm or expected behaviour within a data set. Anomalies, also known as outliers, are data points that differ in some way from most data, making them appear unusual or suspicious.
In this context several AI models have been implemented. Those models are trained on specific manufacturing and quality control data. Preprocessing include cleaning and filtering of anomalous tanks, matching data sources and standarization.
HLEG [102]
HLEG [102])