DC9: Enhancing the dependability of mission-critical system through an advanced anomaly detection methodology for electromagnetic disturbances
Project: Enhancing the dependability of mission-critical system through an advanced anomaly detection methodology for electromagnetic disturbances (WP3)
Host institution: KU Leuven (Belgium)
Supervisor: Mathias Verbeke (KU Leuven, Belgium)
Co-supervisor(s): Davy Pissoort (KU Leuven, Belgium), Dimitar Nikolov (TUS, Bulgaria), Ronny Deseine (Barco, Belgium)
Objectives:
- Create a semi-supervised anomaly-detection methodology tailored to identifying potential EMI.
- Improve the accuracy of anomaly detection in the presence of EMI such that false alarms are minimized.
- Investigate the integration of the EMI-footprint concept into the anomaly-detection methodology.
Expected Results:
- A novel, semi-supervised framework for detecting anomalies in EMI data, which can be applied in real time.
- Machine-learning models that exhibit increased accuracy in identifying EMI while minimizing the number of false alarms.
Planned secondment(s):
- Academic secondment: TUS, Dimitar Nikolov, M15-M18, 3M, Detecting run time anomalies using digital twins.
- Industrial secondment: BARCO, Ronny Deseine, M26-M28, 2M, Applying a run-time anomaly detection method in medical systems
Enrolment in Doctoral degree: KU Leuven (Belgium)