MaaSiveTwin Workshop Showcases AI-Driven Approaches for Rare Earth Element Forecasting
On 21 July 2026, the Technical University of Crete (TUC) hosted the “Data Processing and AI-Driven REE Forecasting” workshop, bringing together leading TUC researchers to explore how artificial intelligence, advanced data processing, and remote sensing technologies can strengthen the monitoring and forecasting of rare earth element (REE) supply chains.
The workshop guided participants through the complete AI workflow for REE forecasting, beginning with the preparation of reliable datasets and progressing to advanced machine learning applications, the broader policy context of critical raw materials (CRMs), and the integration of satellite imagery for mining monitoring.
The first session, delivered by Dr. Marios Antonakakis and PhD student Stelina Naka, introduced an iterative data pre-processing pipeline designed to transform complex, heterogeneous datasets into reliable inputs for AI forecasting models. Through a live Python demonstration, participants gained practical insights into key data cleaning, validation, harmonisation and feature engineering techniques required for robust REE forecasting.
Building on these foundations, Dr. Giorgos Livanos demonstrated how machine learning (ML), deep learning and Earth Observation can support REE supply-chain forecasting. He explored how satellite imagery can be used to monitor mining activity by identifying changes in open-pit mining areas, disturbed land and ground deformation. The presentation also showed how satellite data can be processed and analysed using deep-learning methods to generate useful indicators of mining activity. These insights can complement production, trade, price and disruption data, providing timely information to support REE supply-chain monitoring and forecasting.
The workshop then broadened its focus with a presentation by Professor Konstantinos Komnitsas, who examined the strategic importance of critical raw materials for Europe’s industrial competitiveness and green transition. His presentation explored the objectives of the European Critical Raw Materials Act, ongoing initiatives to strengthen domestic supply chains, and the challenges of translating Europe’s strong research capacity into industrial deployment, fostering an engaging discussion with participants on the future of Europe’s CRM strategy.
In the final session, Dr Nikos Sifakis (Industrial and Digital Innovations Research Group, Technical University of Crete) explored digital shadows, energy systems and circular value chains from a decision-oriented perspective. He explained the differences between digital models, digital shadows and digital twins, highlighting that the value of digitalisation goes beyond simply collecting data or improving forecasting accuracy. Drawing on examples from energy systems and circular value chains, the presentation showed how these tools can support better-informed and more transparent decision-making, while addressing issues such as governance, traceability, environmental impacts and secondary-material supply.
