We’re proud to announce that our research was presented at the 12th IEEE International Conference on Data Science and Advanced Analytics (DSAA) in Birmingham, UK, a leading global forum for interdisciplinary data science research and innovation.
Our researcher Dr. Matej Vukovic showcased the paper titled:
“Leveraging Causal Discovery to Tackle Complexity in Model-based Anomaly Detection: Case-study from Blast Furnace Operation”,
a result of our close collaboration with Primetals Technologies and the BANDAS-Center (University of Graz).
The presented work addresses key challenges in transparency, interpretability and robustness of industrial AI systems. By applying Causal Discovery methods, the research identifies critical process parameters and their temporal dependencies in complex systems such as Blast Furnace operations. These insights are then integrated into predictive models to enhance the understanding and characterization of anomalous behavior, contributing to more trustworthy and explainable AI solutions for industry.
A big thank you goes to the entire research team for their contribution and commitment: Matej Vukovic, Belgin Mutlu, Thomas Kristan, Petra Krahwinkler (Primetals), Christian Tauber (Primetals), and Stefan Thalmann (BANDAS-Center).
