Upcoming TwinODIS Research Talks
Join us on Monday, 28 September 2026, from 13:00 to 14:00 CEST for a hybrid event hosted by ESE EI Econometrics, Erasmus University Rotterdam.
The event will feature two research talks from researchers at FORTH-ICS and the University of Crete, focusing on robust time series analysis, AI-driven forecasting, and smart energy systems.
Matrix Profile Indexing for Multivariate Time Series in Impulsive Environments
Georgia M. Tsanta
University of Crete and FORTH-ICS
Real-world time series from environmental sensors and industrial systems are frequently affected by impulsive noise, where extreme observations occur more often than conventional Gaussian models predict.
In such settings, the Euclidean distance commonly used in matrix profile computation can become unreliable, as large outliers strongly affect similarity measurements.
This talk introduces a generalized matrix profile framework based on an Lp pseudo-norm, enabling more robust motif discovery and anomaly detection under heavy-tailed and impulsive noise.
Experimental evaluation on real-world data shows that the proposed approach can better preserve the ranking of true anomalies compared with the standard Euclidean baseline, particularly as the noise becomes increasingly impulsive.
About the Speaker
Georgia M. Tsanta is a postgraduate student in the Computer Science and Engineering programme at the University of Crete and a research fellow at the Signal Processing Laboratory of the Institute of Computer Science at FORTH. She received her BSc in Computer Science from the University of Crete in 2025. Her research interests include Machine Learning, Data Science, Time Series Analysis, and Artificial Intelligence applications.
Drievly: A Modular Platform for Bridging Price Forecasting and EV Fleet Charging Optimization
Pavlos Charalampidis
FORTH-ICS
This talk will present the software architecture and an early prototype of Drievly, a user-centric modular platform combining AI-driven electricity price forecasting with optimized charging schedules for electric vehicle fleets.
The platform connects energy-market information with fleet operational requirements and charging infrastructure, while supporting interchangeable forecasting models and optimization strategies.
The presentation will conclude with a live demonstration of the platform.
About the Speaker
Pavlos Charalampidis is a Research Software Engineer at the Institute of Computer Science at FORTH. He received his M.Sc. in Computer Science from the University of Crete and his M.Eng. in Electrical and Computer Engineering from the Aristotle University of Thessaloniki.
His research interests include Internet of Things architectures and applications, Smart Cities, IoT security and privacy, cloud infrastructures, cyber-physical systems, and wireless sensor networks.
Join Online
The talks can also be followed remotely through Microsoft Teams.
Meeting ID: 344 728 825 905 375
Passcode: EW9Hk2hm
Join the event via Microsoft Teams
We look forward to welcoming researchers, students, and practitioners interested in time series analysis, artificial intelligence, optimization, smart energy systems, and data-driven decision support.