Safe and high-quality roads: from photo and video recording of accidents - to integrated intelligent road management
Article included in the review TAdviser Intelligent transport system of Russia
Content |
.png)
 | "The driver of ITS development in Russia today is the national project" Safe and high-quality roads "and its subprogramme for creating ITS in agglomerations," says Alexander Semkin, chairman of the GLONASS cluster council. |  |
It is in line with this development that ITS should appear in 64 urban agglomerations of Russia with a population of more than 300 thousand people.
The basis of the national project "Safe and high-quality roads" (BCD) was projects for the introduction of subsystems of photo and video recording of traffic violations and weight and size control. Igor Kravchenko, deputy director for innovative markets at Reksoft, considers photo and video recording systems for traffic violations (FVF) to be the most developed of all ITS subsystems.
| | Today, existing systems increasingly combine a number of functions. For example, the camera can register at the same time the passage to a red traffic light, leaving for the stop line, violation of the speed limit, "says Igor Kravchenko. - The leadership of this class of systems is clear - their introduction pays off with fines, brings constant income to the local budget and equalizes the situation with traffic safety on difficult sections of the road. | |
Targets of the national project "Safe and high-quality highways"
The expert also explains the popularity of the introduction of non-stop weight and size control systems (automated weight and size control system, ASVGK) for financial reasons: "The systems help collect taxes and simultaneously monitor the movement of freight transport."
For example: during the operation of the state system Platon"" (this system of charging for compensation for damage caused to automobile federal public roads by vehicles with a permitted maximum mass of over 12 tons was launched in November 2015), the road fund Russia was replenished by almost 130 billion rubles. According to Ministry of Transport RUSSIAN FEDERATION Rosavtodor, due to additional funding received from the Plato system, more than 2,000 km of problem roads were built and repaired in 40 cities and regions, 700 km of federal routes on sections of 10 roads and 31 bridges selected by freight carriers in 20 constituent entities of the Russian Federation. Rosavtodor sends all revenues from 2020 to 2022 from the operation of Plato to overhaul more than 100 bridge structures and 600 km of loaded federal routes.
Weight and size control: there is a lot of work ahead
| | The relevance of weight and size control has increased due to the development and improvement of transport. The weight of the cargo that can be carried on one vehicle has increased significantly, but the roads were not ready for this, so the criticality of the issue of preserving the roadway has increased significantly. | |
According to Softline forecasts, in 2022 at least 30% more projects will be implemented, provided that there are no force majeure and a sequester of the budget.
Indeed, the expert notes, according to the draft BCD, each region has KPI for FVF and ASVGK systems. After all, they fulfill an important social and financial mission: they improve road safety and bring indirect economic effect - the region spends less money on road restoration and can direct efforts to develop a regional transport network. So, in the Omsk region alone, Softline is developing the ASVGK system: 9 of 16 intelligent weight and size control systems are already operating on the roads of the region, another 7 are in the process of installation and integration into common systems.
| | Automated weight and size control is a technical solution that each region needs. First of all, because such a system allows the subject to increase traffic safety and preserve roads, "said Nikolai Ivanenko. - In addition, there is already a trend towards the preparation and development of the road network from the point of view of unmanned vehicles. I am confident that this trend will remain current and reach its peak over the next five years. | |
For example, the Irkutsk region will allocate 10.6 billion rubles this year for overhaul, construction and reconstruction of roads, says the governor of the Irkutsk region Igor Kobzev, but how to make sure of the quality of the repairs? It was decided that the quality of repaired roads will be checked using laser scanning technologies. An even level of coverage, compliance with slopes and other geometric parameters of roads will be evaluated by a special mobile laboratory directly during the movement. This means the possibility of quickly conducting high-precision surveys of a sufficiently large volume in a short period of time without blocking the road section and attracting additional specialists, said Maxim Lobanov, Minister of Transport and Road Facilities of the Irkutsk Region.
Overloaded vehicles continue to spoil Russian public roads, noted with regret at the Russian Academy of Transport. According to industry experts, the amount of annual damage caused to public roads by congested vehicles amounted to 2.6 trillion rubles. At the same time, every sixth heavy truck moves along federal routes with a significant margin (an average of 41%), says Andrei Samaryanov, deputy head of Rosavtodor. He cited these data at a joint meeting of public councils at Rosavtodor and Rostransnadzor, which took place in mid-February 2022 and was specifically devoted to road safety issues.
According to Andrei Samaryanov, only in 2021 the lost amount of administrative fines for violations by total mass could amount to about 200 billion rubles,
| | And this is only 12 points of weight and size control, "the official emphasized. - The share of violations in terms of weight parameters for a separate category of vehicles, for example, a five-axle road train, reaches 25%. The same category and the most massive is about 70% of the total number of all violators. | |
Participants in the event unanimously advocated toughening the fight against vehicle outliers on federal and regional roads. So, according to Vladimir Belozerov, chairman of the public council under Rostransnadzor, according to previous plans, it was planned to install 387 VGK points on federal highways by 2027, and on the network of roads of regional importance in 75 constituent entities of the Russian Federation - 366 points by the end of 2024. These plans have been revised in the greater direction, Igor Starygin, director of the Russian Association of Territorial Highway Management Bodies, noted: local authorities agreed to increase the network of automated VGK points to 597. As a result, the total number of subjects of the Russian Federation covered by the system will increase to 81. To date, according to RADOR, 270 APVGK have already been installed in 52 regions.
He also explained that it is impossible to block the branched regional network by automatic points of the Supreme Command alone, since there are always options for bypassing alternative routes. In this regard, it is necessary to expand the network of not only automatic, but also stationary and mobile VGK points, Igor Starygin is sure.
A little earlier, in January 2022, the government commission on legislative activities considered an initiative that strengthens control over heavy and large trucks. According to its decision, an automatic system is introduced to ensure that trucks do not exceed the standards for weight and size. Special complexes will fix such violations. The prepared amendments to the Administrative Code empower Rostransnadzor to consider cases of violations of traffic rules of a heavy or large vehicle.
How recognition technologies are evolving
The stationary all-in-one camera RNCam of the company "Recognition Technologies" combines in a single case a video sensor and a protected industrial computer with the functionality of photo video recording of traffic violations, measuring speed and searching for vehicles. At the same time, one RNCam camera monitors up to four lanes and can support the operation of two additional viewing cameras, actually replacing the error of 1 km/h, more than 100 km/h and the error of 2 km/h with the usual one when measuring speed up to 100 km/h. And the speed of vehicles, which can reach 255 km/h, is measured radar-free - by video frames with an error of up to 2 km/h.
RNCam Fixed All-in-One Camera Architecture
To recognize car numbers, RNCam uses the new AvtoUragan-VSM2 algorithm. It ensures the operability of the complex, even if the chamber is located at an acute angle with respect to the moving traffic flow.
In the stationary version - a complex for automatically recognizing state registration marks of vehicles and fixing violations of the AvtoUragan-VSM2 traffic rules - these technologies solve the problems of identifying, registering and controlling the movement of vehicles in a stream of any intensity: from small car parks to multi-lane highways. It provides the recognition of license plates in daylight at 97%, as well as a high degree of recognition of dirty or badly damaged license plates.
The operating system is not fixed by radar detectors. Not only exceeding the speed limit, but also a number of other traffic violations are recorded, as well as automatic search for vehicles from databases is provided.
ICL has developed a number of specialized machine vision and video analytics solutions designed for the transport industry: monitoring the condition of hatches on the roadway (identification of open hatches, their identification according to the road passport, sending a message), search and classification of defects on the road (monitoring the roadway in order to detect pits, cracks and other defects), search for non-functioning lighting poles, monitoring road signs (monitoring the state of road signs or their absence). Vladimir Kayurov, senior project manager of the ICL group of companies, says that there are a lot of such types of objects on the roads, and therefore it is necessary to use a non-trivial system for determining their geographical positioning. He emphasizes that any video material can be used for the operation of the corresponding system, for example, recorded on a car recorder or even shot by a smartphone camera.
How recognition technologies are smarter
The flagship direction of the development of FVF technologies is the expansion of the set of recognized situations using mathematical methods of neural networks. For example, the mobile hardware and software complex (AIC) "ParkRight" of the company "Recognition Technologies," designed to fix traffic violations from the interior of any car, can record violations of parking and stopping rules, the passage of a pedestrian crossing, route transport lanes, oncoming traffic lanes and other types of violations, for which, as an evidence base, a photo of a car with a recognized state registration mark, a photo and a video from a viewing camera are sufficient. If necessary, this video can be framed and printed.
The complex automatically checks all cars that fall into the visibility zone of the cameras against connected search bases. When a wanted car is detected, the operator receives a visual and audible alert, and the data obtained during the car patrol is automatically transmitted via wireless communication channels to the data center.
The mobile AIC for photo and video recording of violations of Avtonom traffic rules of Recognition Technologies the company, equipped with a geo-positioning receiver/, GLONASSGPS is installed on the side of the highway at an arbitrary angle to the roadway and automatically registers all vehicles entering the control zone. Moreover, the use of multi-core processor allows the Avtonom complex to determine the speed of multi-directional movement, both oncoming and outgoing traffic.
In addition to recognizing the license plate and the speed of the vehicle at the time of its fixation, Avtonom is able to detect such traffic violations as driving along the side of the road, driving along the oncoming lane, driving along the public transport lane, violation of the row of traffic on the lanes.
Last summer, AvtoUragan-VSM2 systems learned to identify drivers who control vehicles without a fastened seat belt in automatic mode.
The Moscow Center for Traffic Management has planned the next stage in the development of an intelligent recognition system: in addition to the functionality of detecting accidents on city roads, it will record garbage, the appearance of animals or stopped cars, as well as fires and other emergency situations on the roads. The data center clarifies that the new devices will not issue fines for the offense, but will send information about incidents to the police and utilities on a real scale, helping emergency services quickly respond to any emergency.
The project worth 4.2 billion rubles. involves the installation over five years of 1.3 thousand video cameras on the Moscow Ring Road and other metropolitan highways.
Speaking about the potential of neural networks, do not forget that the most important thing in the qualitative work of the ML algorithm is the data on which the algorithm was trained: it is the work with data, including the preparation of a data set for training the neural network and validating the operation of the algorithm, that is critical for the operability of an intelligent solution. In this part there is space for competition of smart recognition solutions.
| | Working with data provides us with the ability to effectively train "light" models and obtain object detection and classification accuracy comparable to deep neural networks (more than 95%), and we constantly improve both the data set and its preparation process. | |
The ability to use "light" neural networks allows, in particular, to run recognition software on low-power single-board computers based on CPU. The lack of GPU requirements greatly enhances the application of Intelligent Video Analytics solutions. For example, it allows you to follow the advanced concepts of IoT, Edge Computing, as well as Edge AI and implement autonomous scenarios for detecting, analyzing and optimizing traffic flows without the need to transfer a video stream or other large amounts of data for processing to data centers over the Internet.
| | Reducing restrictions on infrastructure and Internet channel opens up the possibility of implementing time-delay-sensitive solutions, such as, for example, adaptive control of connected traffic lights in real time, where the delay for deciding to change the traffic lights should not exceed 5 seconds, "emphasizes Vyacheslav Lukin. And he adds: Over the past few years, ITS technologies have significantly stepped forward due to such phenomena as big data (Big Data), artificial intelligence (AI) and [[Internet of Things (IoT)|Internet of things of Internet of Things (IoT)]]]]]]]]]]]]]]]]]]]]]. All these technologies allow you to implement scenarios that previously could not even come to a head. Although these technologies are not always mature enough to be introduced in "combat" projects, they significantly advance ITS technology as a whole. | |
According to the expert, computer vision technologies, although dynamically developing in the field of ITS, still gravitate towards decisions related to photo and video recording or fines: detection of a strapped belt, a phone in the hands of the driver, turned off headlights, etc.
| | This is probably the most transparent story for the state customer, "Vyacheslav Lukin believes. - It is quite clear how to assess the costs and effect of the introduction in money in order to evaluate the ROI of such decisions. If we are talking, for example, about optimizing traffic flows, then estimating the effect of such decisions is not so easy. I think that this is one of the stop factors for the development of complex and socially significant, but not so obvious in assessing the effect of ITS. | |
Enhance VF capabilities through integration with other systems
| | Today, the main areas of activity of ICL in the field of ITS are the introduction of automatic points of weight and size control, the development of video analytics systems based on neural networks and the development of a Unified platform for managing transport systems, the expert notes. | |
An example of the implementation of such an idea is the opening last November of the first section of the Central Ring Road (CKAD). Today it is the most technology-rich motorway in Russia, experts say. And the non-stop fare system, the intelligent traffic management system of the new generation and the unified toll collection system are unique for the Russian market. Information from the AvtoUragan-VSM2-M FVF road complexes is used to determine the cost for the actually traveled section of the TsKAD toll highway, which vehicles pass without reducing speed. The Free Stream solution automatically classifies cars, recognizes their license plates and bills them, which can be paid through a mobile application, a personal account on the site or using a transponder. Thus, the first highway with a barrier-free fare payment system appeared in Russia.
In addition, the new generation automatic traffic control system introduced at the Central Ring Road System collects information from FVF complexes on weather conditions, accidents, traffic density and congestion, as well as on the appearance of foreign objects, pedestrians or animals on the road.
MVS Group, which implements the largest concession in the field of road safety in Russia (more than 2,200 FVF complexes in operation), has developed a special solution - the "BDD Analytical Module," which is an intelligent software for analyzing and predicting emergency sections of roads and accident centers. Data aggregation functions (collection of all types of data from different sources for accounting and analysis of accidents), statistical processing (detailed information on statistics of violations recorded by FVF systems) and forecasting are supported. Predictive analytics methods help predict the occurrence of accidents and offer optimal options for preventing them, including changing the organization of traffic, correcting and controlling speed, installing traffic lights and artificial irregularities.
In fact, this is a single information environment for all participants in road safety, they say at MVS Group: all data from regional and federal systems are collected into a common base in a single format, which eliminates the discrepancy in indicators and ensures the speed and quality of interaction between all parties.
A project of this class was implemented in 2020 in the Penza region. The region's automated road safety system was created by order of the Safe Region State Budgetary Institution by AIT Software specialists based on the Forsyth software product . Analytical Platform. " The platform developer is Forsyth.
The main goal of implementing the system is to reduce the number and severity of the consequences of accidents thanks to a deep automated analysis of their causes. The system collects and constantly updates data on the state of roads, as well as information on accidents in all areas of the Penza region. This allows you to quickly identify the places of concentration of accidents according to given parameters: time period, territory, types of accidents, and visualize the situation with accidents in the form of various diagrams and tables with representations of accident indicators and places of concentration of accidents for comparative analysis of the same type of territories and time intervals.
Accident centers on the section of the Federal Road M-5 in 2019-2020
On the electronic map, road signs, road arrangement components and places of accident concentration related to different periods of time are indicated to analyze the effectiveness of implemented measures aimed at reducing the number and severity of accident consequences.
Based on the analysis of data, various groups of measures are formed aimed at reducing the number and severity of the consequences of accidents, including organizational, information, educational and educational, engineering, construction and road, operational.
Example of data generation for road activity planning
According to the estimates of Alexander Markin, head of the program support department of the Situation Center of the Governor of the Penza Region, as a result of the work of the system in the Penza Region in the first 9 months of 2021, compared to the same period last year, a reduction in economic damage from accidents in the amount of more than 850 million rubles was achieved, a decrease in the number of accidents with victims by (5.31%), the number of deaths - by 22.14%.
| | Urban automated traffic management systems (ASUDD) are becoming increasingly intelligent. They focus on the intensity of traffic flow and try to solve the problems of optimizing the operation of traffic lights. The main ASUDD emphasizes informing drivers using a scoreboard. They show information about the permitted speed, road works, traffic jams in front. Often, such ASUDDs are associated with weather stations and give dynamic information about the state of the road surface, contain the function of tracking traffic, recognizing the types of vehicles in the stream, etc., "summarizes Igor Kravchenko, deputy director for innovative markets at Reksoft. | |
How do I create complex solutions of this class?
Other review materials
- Intelligent transport system of Russia: course - on a national scale
- Specifics of ITS
- Urban and regional ITS
- Road Traffic Management Best Practices
- Public transport management
- Digital Logistics
- Advanced telematic solutions and smart cargo delivery
- ITS infrastructure. Platform and service approach
- Intelligent Transport System Services
- Establishment of a national ITS
- Digital road
- Promising development of ITS: unmanned transport
| The site content is translated by machine translation software powered by PROMT. The machine-translated articles are not always perfect and may contain errors in vocabulary, syntax or grammar. Read original article If you find inaccuracies or errors in the results of machine translation, please write to editor@tadviser.ru. We will make every effort to correct them as soon as possible. |