DigiPave: Digital-Twin-Enabled and Data-Driven Pavement Maintenance Planning

DigiPave

DigiPave develops an advanced digital twin platform for fully automated, data-driven road maintenance. Combining real-time IoT monitoring, high-fidelity virtual modelling, and state-of-the-art AI (YOLO, GANs, VAEs, LLMs, JEPA), it delivers crack segmentation, defect detection, pavement condition assessment, crack evolution prediction, and automated PCI/SCI calculation.
Digipave_resized
Digipave project simple introduction figure

Project information

Project duration

-

Funded by

Business Finland

Funding amount

1 214 038 EUR

Project coordinator

University of Oulu

Contact information

Project leader

Contact person

Other persons

Project description

The project develops a next‑generation digital twin platform for road pavements, enabling fully automated, data‑driven maintenance planning. Building on advanced AI, IoT, and edge-cloud technologies, the project integrates real‑time pavement monitoring with high‑fidelity virtual modelling to create a continuously updated digital representation of the road network. Within this platform, we have developed a suite of state‑of‑the‑art AI pipelines capable of performing complex downstream tasks, including crack segmentation, crack classification, defect detection, and pavement condition assessment. These models leverage cutting‑edge architectures such as generative models (including GANs, VAEs, LLMs), YOLO, and Joint Embedding Predictive Architectures (JEPA) to identify and characterize defects like linear cracks, alligator cracking, potholes, and surface deformation with high accuracy and minimal human intervention.
The DigiPave platform goes beyond detection: it enables crack evolution prediction, automated computation of Pavement Condition Index (PCI) and Surface Condition Index (SCI) and ultimately supports optimized Maintenance and Repair (MnR) decision‑making. By automating the entire workflow from sensing to prediction to planning, the project significantly reduces operational costs, enhances maintenance timing, and improves long‑term pavement performance. Overall, DigiPave demonstrates how digital twins and AI can transform road asset management into a proactive, scalable, and resource‑efficient process.
The project is funded by Business Finland, with the University of Oulu as the main research partner, supported by Volvo, Unikie, Satel, Infrakit, and the Finnish Meteorological Institute.

Project results

List of authors Title Venue Link
Abdelhak kharbouch, Mehdi Rasti, Amirhossein Ayoubi Khosroshahi Demo: 6G‑enabled digital twins for multi-domain critical infrastructure monitoring EUCNC26, Malaga, Spain NA
Mehdi Rasti, Abdelhak kharbouch T8- 6G-Enabled Digital Twins for Verticals: Architecture, Platforms, and Use-Cases PIRMC 2025 Istanbul, Turkey https://www.6gflagship.com/event/pimrc2025/
Shiva Kazemi Taskou , Mehdi Rasti , and Ekram Hossain End-to-End Resource Slicing for Coexistence of eMBB and URLLC Services in 5G-Advanced/6G Networks IEEE Transactions on Mobile Computing https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10354514
Daneshvar Arya, Golroo Amir, Rasti Mehdi Incorporating physics-informed neural networks into yolo for pavement rutting detection AM3P https://doi.org/10.34726/10885
Entezari Mohammad Saleh, Golroo Amir, Rasti Mehdi A New Era in Transportation Infrastructure Contrition Evaluation with Connected Vehicles AM3P https://doi.org/10.34726/10578
Gholipoor Narges, Kharbouch Abdelhak, Rasti Mehdi, Golroo Amir Optimizing Storage for Real-Time Road Monitoring with Dual-Camera Systems AM3P https://doi.org/10.34726/10575
Najafli Mohammad Amin, Fahmani Mohammad Sadegh, Golroo Amir, Rasti Mehdi Enhancing pavement performance evaluation via crowdsourced Gamification AM3P https://doi.org/10.34726/10604
Sedighian-Fard Mohammad, Golroo Amir, Nouri Salar, Rasti Mehdi Autonomous synthetic data generation for asphalt pavement crack segmentation using generative models AM3P https://doi.org/10.34726/10781
Talaghat, Mohammad Amin, Sedighian-Fard Mohammad, Golroo Amir, Rasti Mehdi Leveraging Digital Twin for Data-Driven Pavement Maintenance AM3P https://doi.org/10.34726/10886
Talaghat Mohammad Amin, Golroo Amir, Rasti Mehdi Improving pavement distress segmentation with diffusion-based generative AI and digital twin insights AM3P https://doi.org/10.34726/10798
Mohammad Amin Talaghat, Amir Golroo, Abdelhak Kharbouch, Mehdi Rasti, Rauno Heikkilä, Risto Jurva Digital twin technology for road pavement Automation in Construction https://doi.org/10.1016/j.autcon.2024.105826
Jianqi Zhang, Ling Ding, Wei Wang, Hainian Wang, Ioannis Brilakis, Diana Davletshina, Rauno Heikkilä, Xu Yang Crack segmentation-guided measurement with lightweight distillation network on edge device Computer-Aided Civil and Infrastructure Engineering https://doi.org/10.1111/mice.13446
Shiva Kazemi Taskou , Mehdi Rasti , and Ekram Hossain Energy-Efficient Resource Allocation for FeMBB and eURLLC Coexistence in RSMA-Based Wireless Networks IEEE Transactions on Cognitive Communications and Networking https://doi.org/10.1109/TCCN.2025.3578509
Narges Gholipoor; Mehdi Rasti; Fahimeh Aghaei; Farid Hamzeh Aghdam; Abdelhak Kharbouch; Valiollah Talaeizadeh A Review on the Cross-Sector Resource Management Framework for Electric Vehicles Integration: Challenges, Solutions, Key-Enabling Technologies, and Future Directions IEEE Open Journal of Intelligent Transportation Systems https://doi.org/10.1109/OJITS.2025.3593437
Mohammad Amin Talaghat,Amir Golroo,Vahid Shahhosseini & Mehdi Rasti Enhancing pavement distress segmentation model with diffusion-based generative AI: a digital twin perspective International Journal of Pavement Engineering https://doi.org/10.1080/10298436.2025.2540076
Mohammad Sedighian-Fard, Amir Golroo,
Mahdi Javanmardi, Alexandre Alahi, Mehdi Rasti
Data generation for asphalt pavement evaluation: Deep learning-based insights from generative models Case Studies in Construction Materials https://doi.org/10.1016/j.cscm.2025.e05116
Mehdi Monemi, Maryam Chinipardaz, Mehdi Rasti, Mehdi Bennis, Matti Latva-Aho Tutorial on Joint Embedding Predictive Architectures (JEPA): Foundations, Applications, and Future Directions ACM Computing Surveys https://openreview.net/pdf?id=Zr4PUe0ZNl
Mohammad Amin Talaghat, Amir Golroo, Mehdi Rasti and Vahid Shahhosseini Road Pavement Digital Twin: A Blueprint with a Case Study Journal of Computing in Civil Engineering https://doi.org/10.1061/JCCEE5.CPENG-7187
Soroush Amiri; Fereidoon Moghadas Nejad; Mehdi Rasti; Mehdi Monemi; Hannaneh Dehghan Tezerjani A Novel Approach to Pavement Crack Classification using Joint-Embedding Predictive Architectures Automation in Construction NA
I. Niskanen, J. Ikonen, M. Immonen, R. Heikkilä Infrastructure-Mounted 2D LiDAR System for Vehicle Detection and Speed Estimation on Motorways IEEE Sensor NA