DC12 Project: Sustainable Real-time Anomaly Detection for Practical Applications

Doctoral Candidate

Athanasia Kollarou, MSc 

Athanasia Kollarou received her Master’s degree in Digital Systems Security from University of Piraeus, with background in Comparative Analysis of Adversarial Attacks on AI Models. Her research expertise spans areas like network security and information security, cryptography, threat modelling, closely aligning with the TUAI project’s goals.

Athanasia_Kollarou_CV.pdf 

Main Supervisor: Jia-Chun Lin (NTNU)

Co-Supervisors: Sandeep Pirbhulal (NRS), Victor Rodriguez-Fernandez (UPM), Volker Stolz (HVL), Dariusz Mrozek (SUT) 

R&D cooperation: AIUT

Objectives: explore approaches for anomaly detection that operate under real-time constraints while remaining energy- and resource-aware in safety-critical cyber-physical systems. 

The focus is on balancing accuracy, latency, privacy, robustness, and interpretability so that timely, reliable insights can be delivered where data is produced and decisions are made.

Research challenges include meeting real-time deadlines with low energy and resource use while keeping detection accurate. Further challenges are resilience to sensor noise and faults, integrating privacy-preserving techniques, and defending against adversarial attacks such as spoofing and data poisoning.

Expected Results: include a concise overview of requirements and sustainability considerations for anomaly detection solutions in safety-critical systems. Additionally, outlining and evaluating a proposed approach that balances accuracy, latency, energy use, privacy, and robustness for timely detection.  

Applied research: The research result of this PhD project will be applied to the cyber-physical electricity systems in NRS’s the National Smart Grid Laboratory.  DC12 will also participate in AIUT’s applied projects on smart manufacturing.

Planned secondments: UPM(4 months); HVL(4 months); SUT (4 months)                   

Enrolment in Doctoral degree: NTNU