This Science News Wire page contains a press release issued by an organization and is provided to you "as is" with little or no review from Science X staff.

Russian Scientists Train Neural Networks to Monitor Traffic

May 12th, 2020 South Ural State University
Russian Scientists Train Neural Networks to Monitor Traffic
Artificial intelligence might help control city traffic. The network was taught by scientists at South Ural State University. They developed an intelligent system for monitoring traffic in real-time. They obtained a patent for a unique technology; now it is being prepared for a "pilot" launch.

The Patent for Unique Design

Car traffic in large cities in Russia is increasing, and it is becoming really difficult to control it. To solve this problem, programs that collect information from highways have been created. However, the delay in data transfer was up to 15 minutes. Scientists of the South Ural State University have developed a unique artificial intelligence system of traffic monitoring. This is part of the large project "Smart City", which involves the use of information and communication technologies to control the functioning and development of the metropolis.

The program collects and sends information about the situation on the roads in real-time. Up to 400 parameters are analyzed at the same time, and the data error is less than 10%. This was achieved through the use of trained neural networks. The work does not require large costs for hardware and video cameras.

Professors, graduate students, and students from the Department of Automobile Transport, the Department of Applied Mathematics and Programming, and System Programming of South Ural State University are involved in creating a promising development. Now a patent for the software has been received. They developed an intelligent system for monitoring road traffic and transport infrastructure.

"Now, our group is developing methods and algorithms for processing big data in the tasks of creating a unified decision-making system for the entire transport system of the city based on artificial intelligence. The developed system, based on the integration of modern information and telematic technologies, is intended for the automated search and adoption of the most effective traffic management scenarios. The collection and processing of big data in real-time allow the Intelligent Transport System (ITS) to instantly assess the current state, signal incidents, predict the development of events and make management decisions," the head of the Smart Transport project, associate professor of the Automotive Transport Department of the Polytechnic Institute of SUSU Vladimir Shepelev said.

The patented technology will be commercialized. The monitoring system was tested at one of the crossroads in Tyumen in 2019. The Department of Transport of St. Petersburg is also interested in an intelligent system which makes it possible to predict traffic jams, to collect and process data on exhaust gases emitted by motor vehicles.

Russian Scientists Train Neural Networks to Monitor Traffic
Credit: SUSU

System improvement continues

The intelligent monitoring system for road transport is being improved, and all scientific developments are recorded in highly rated scientific journals. The latest patent technology article was published in Transport and Communication (Q2). Scientists proposed a new system for analyzing the situation on the roads in order to solve the problem of traffic jams.

The methods provide accurate detection of road users: not only large but also small vehicles and pedestrians are recorded, which affects the traffic. The developed solution is more accurate and the speed of information transfer is higher.

"We have developed a data collection system for detecting vehicles while monitoring urban traffic. It is a software based on neural networks. To train the system, a data set was collected in which for each crossroads there were about 1000 images taken at different times of the day under different weather conditions. This allowed us to receive data in any condition. Statistical analysis of the collected information using factorial, cluster, regression and multidimensional scaling methods made it possible to identify the most important intersection characteristics that affect their traffic capacity under congestion conditions. The analysis helped to make forecasts depending on the initial parameters of roads, provided that segmentation of roads by initial characteristics and visualization of the results are realized," the authors of the article explained.

A team of scientists from SUSU is planning further research. Specialists intend to use data from touch-sensitive traffic cameras in their work to determine traffic with traffic jams and without them. Such studies will improve the road transport infrastructure of urban networks. That is precisely the purpose of using an intelligent monitoring system.

Significant results have been achieved in the development of Information technologies at South Ural State University. Supercomputer modeling is being used in research, processing, storage and mining of Big Data.

IT is one of three strategic areas for the development of scientific and educational activities of South Ural State University along with ecology and materials science.

SUSU participates in the 5-100 Project intended to increase the competitiveness of Russian universities among the world's leading research and educational centers.

More information: … _of_the_Road_Network

Provided by South Ural State University

Citation: Russian Scientists Train Neural Networks to Monitor Traffic (2020, May 12) retrieved 1 March 2021 from
This document is subject to copyright. Apart from any fair dealing for the purpose of private study or research, no part may be reproduced without the written permission. The content is provided for information purposes only.