Contextual anomaly detection, complex & heterogenous data
The growing amount of data that is continuously gathered by industrial machinery and processes is increasingly being exploited for the operational optimisation of these industrial assets. A cross-cutting challenge in this data-driven optimisation process is the identification of rare events or observations that deviate from normal behaviour. Typically, this anomalous behaviour is connected to a problem with the industrial equipment, such as structural defects, malfunctioning equipment, etc., indicating that the asset is no longer operating in its optimal state. The highly dynamic conditions in which these assets operate make the identification of this anomalous behavior a major challenge.
This challenge will be addressed in the CONSCIOUS project, where the overall objective is to research effective solutions to achieve a more accurate, robust, timely and interpretable anomaly detection in complex, heterogenous data from industrial assets by accounting for confounding contextual factors. The results will be validated on multiple real-world use cases in different domains.
January 2021 - June 2023
With the support of (met de steun van)