The scientific team of Novosibirsk State Technical University (NETI) has developed a solution that allows automating the allocation of structural elements of the roadway. The created software product is planned to be used for designing and certifying roads and using it to solve problems related to the development of the region's automotive transport infrastructure. The Ministry of Transport and Road Management of the Novosibirsk region highly appreciated the development of the university.
In November 2024, NSTU-NETI and the Territorial Administration of Highways of the Novosibirsk Region launched a project to develop software for creating digital models of highways that will help determine geometric parameters, take into account bumps, stops, road signs and other elements. During the first stage of the project, specialists from the Department of Theoretical and Applied Computer Science have developed an application that combines elements of geographic information systems (GIS) and computer-aided design (CAD) systems.
"The Novosibirsk region has a long network of highways, and it is important for us that the work carried out is carried out efficiently, including at the design stage. The development of NSTU-NETI makes it possible to automate the collection and analysis of road parameters, reduce the influence of the human factor and improve the accuracy of design and passport work. Such digital solutions work directly to improve the quality of the region's road infrastructure, and they have not only scientific but also practical potential," said Anatoly Kostylevsky, Minister of Transport and Road Management of the Novosibirsk Region.
"Currently, laser scanning is one of the modern, actively developing and applied technologies for spatial data collection. After receiving the initial data, specialized programs are required for their further processing — filtering, thinning, and highlighting the structural elements of the roadway. An information search for analogues and their analysis showed that existing solutions do not always pay enough attention to design elements, and often design specialists need to export data to other programs for further processing. We decided to create our own product — a comprehensive solution that can be called "GIS with CAD elements", which allows us to optimize the functionality and increase the productivity of work on the formation of a road passport," said Vitaly Karmanov, project manager, Associate Professor of the Department of Theoretical and Applied Informatics, Candidate of Technical Sciences.
According to one of the team members, Andrey Krivetsky, an engineer at the Info-Mir Research and Educational Center of NSTU-NETI, before creating the software, the developers proposed and researched machine learning methods for automatic detection and detection - highlighting the boundaries of roadway elements (markings, dividing strips, pavement edges, slopes, etc. etc.).
Currently, the research team has started the second stage of the project. By November 2026, the program should be supplemented with the ability to determine the geometric parameters of the longitudinal and transverse profile of the road, to identify defects in the roadway. It is planned to use Big Data technologies, machine learning elements, computer vision, artificial intelligence theory and neural networks to develop mathematical, algorithmic and software for the project.