AI to Safeguard Road Quality
Researchers at Donskoy gosudarstvennyy tekhnicheskiy universitet – DGTU (Don State Technical University) in Rostov-on-Don are developing an artificial intelligence algorithm designed to assess road conditions and generate optimal repair schedules. The goal is to reduce unnecessary spending while keeping road infrastructure in better condition and making maintenance decisions more data-driven.
AI-Powered Maintenance Planning
Scientists at DGTU are developing an AI algorithm that analyzes long-term pavement diagnostics, predicts how road conditions will evolve and ranks repair priorities. The central concept is to move away from fixed maintenance intervals – typically every 12 to 24 years – and instead schedule repairs based on the actual condition of the road.
This approach could reshape how road infrastructure is managed. For drivers and local communities it means smoother roads, faster repairs on deteriorating sections and improved traffic safety. For governments it offers more efficient use of public funds and a shift toward digital lifecycle management of infrastructure. The development also strengthens Russia’s position in building domestic AI technologies for the transport sector.
Where the Algorithm Could Be Used
In the long term the technology could become part of the broader digitalization of Russia’s transport infrastructure. The algorithm can integrate with regional road management systems, generating dynamic maintenance plans and forecasting pavement wear using data collected by mobile laboratories, cameras, sensors and diagnostic systems.
The system could be deployed on federal highways, regional roads and urban street networks – anywhere infrastructure condition requires accurate and timely monitoring.
The project also carries export potential. Similar systems are in demand across the CIS and in countries with rapidly developing transport networks. The technology also fits into Smart City initiatives, where AI-driven road monitoring could support international projects aimed at improving urban mobility and infrastructure management.
Toward Smarter Roads
The DGTU project is not an isolated experiment but part of a global trend toward using AI to manage transport infrastructure. In recent years both Russia and other countries have launched initiatives exploring similar technologies.
In 2023 Moscow began using neural networks to automatically detect road defects such as potholes, damaged curbs and worn lane markings. Mobile monitoring systems collect the data while AI analyzes the images and identifies problem areas.
Between 2021 and 2024 several Russian research programs studied the use of machine learning to forecast traffic flows and road conditions. At the same time international teams have advanced computer vision technologies capable of detecting cracks and structural damage in pavement images.
These examples show that the Rostov-based development builds on a growing body of research and addresses real operational challenges faced by road authorities.
Roads of the Future
The algorithm being developed at DGTU illustrates how AI tools can reshape road infrastructure management. Moving toward neural network-driven infrastructure management based on data collected by sensors and connected devices could reduce maintenance costs while increasing the efficiency of public investment. Over time the technology could become part of a unified national system for monitoring Russia’s transport infrastructure integrated with Smart City platforms.
Within the next three to five years similar systems are expected to scale up from pilot deployments in individual regions to nationwide digital platforms for road management. If the DGTU solution proves effective it could also enter international markets as a software platform for infrastructure condition analysis. In that scenario the Rostov-developed AI could help pave the way for safer and more reliable roads not only in Russia but worldwide.
