The IVORY doctoral candidates showcased their research at the 12th International Congress on Transportation Research (ICTR 2025), held from October 16–18, 2025, in Thessaloniki, Greece.

The conference brought together academics, researchers, and professionals from across Europe to exchange ideas and innovations shaping the future of transportation systems, AI applications, and road safety. Several IVORY researchers presented their latest findings, demonstrating the project’s interdisciplinary approach and its growing contribution to data-driven, AI-enabled road safety research.


🔹 Simone Paradiso (DC9): “Aggregating Telematics for Road Safety Analysis

Simone presented a refined method for aggregating telematics data onto road networks, introducing a structured framework that integrates statistical and machine learning techniques. This approach aims to improve accuracy beyond traditional aggregation methods. The work was conducted in collaboration with OSeven Telematics, Simone’s industrial partner within the IVORY Network.

👉  The final paper can be seen here.


🔹Aristotelis Tsoutsanis (DC7): “Identifying Dangerous Street Segments and Traffic Behavior in Athens Using Telematics, Crash and Traffic Data

Aristotelis presented his work in Session S2B, focusing on identifying hazardous road segments in Athens by combining telematics, crash, and traffic data. The analysis provides valuable insights into urban safety assessment and supports data-driven policy interventions.

👉  The final paper can be seen here.


🔹Amirhossein Hassani (DC3) & Muhammad Shahid (DC11): “EfficientNet-Swin Transformer for Automated iRAP Road Safety Attribute Extraction in Low- and Middle-Income Countries

Amirhossein and Muhammad presented a hybrid EfficientNet-Swin Transformer model for multi-task classification of seven iRAP road safety attributes using image data from Kenya. Their work tackles key challenges such as class imbalance and label inconsistency in LMIC datasets, with future research focusing on temporal context and synthetic data generation for rare classes.

👉  The final paper can be seen here.


🔹Júlia A. Porto (DC8): “Road Segmentation Made Simple: A Practical Comparison of Segmentation Models and Post-Processing Techniques

Júlia presented her early-stage research, conducted with the NTUA team, offering a practical comparison of lightweight segmentation models and graph-based post-processing techniques for aerial imagery. Her study discussed multilevel data integration, computational challenges, and GDPR compliance in transportation applications, as part of Session S2C, Big Data for Transportation.

👉  The final paper can be seen here.


The active participation of IVORY PhD researchers at ICTR 2025 underlines the network’s commitment to advancing AI-based road safety, data analytics, and sustainable mobility across Europe. Congratulations to all participants for their impactful contributions!