The Urbantech Group brings together developers and integrators of smart mobility solutions; the Group also develops its own technological solutions for the transportation market. One of our ecosystem companies, MVS Group, the operator of the traffic enforcement camera system in the Moscow Region, operates more than 2,250 TEC units. Based on infrastructure created earlier, several solutions have been developed that we are currently rolling out in more than 10 Russian regions. Our team focuses on the main dimensions of the Russian Transport Ministry’s transportation strategy, i.e. digital twins of road infrastructure, digital management of Russia’s national transportation system, digitisation in the field of transport security. We develop solutions for road infrastructure planning and modelling, traffic management, safe traffic arrangements, and public transport upgrades.
Is there a distinct push for transport planning and modelling in Russia?
In Russia, around 60 billion roubles are spent every year on preparing various plans intended to support the development of transport infrastructure and end-to-end mobility schemes. However, only 10-20% of the proposed projects are brought to life. The main causes are the insufficiency of up-to-date data for the formulation of transport infrastructure development plans and end-to-end mobility schemes, lean funding, non-digital data collection and processing formats. Another obstacle is a commonplace approach where various regional level services and departments responsible for the implementation of smart mobility systems, public transport upgrades and road network reconstruction fail to synchronise their infrastructure development plans.
To resolve transport planning issues, it is important to analyse the current mobility level of a city and predict how this will change as its street and road network evolves. It is no less important to use up-to-date (and continuously updated) data on the condition and location of road infrastructure facilities: traffic signs, traffic lights, pedestrian crossings, road markings and public transport stops. A combination of these approaches will allow for the creation of a fully-fledged digital twin of the city, the formulation of transport infrastructure development strategic plans and the accomplishment of lower-level tasks: SMS implementation, road safety improvement, traffic management and combatting traffic jams.
What is your proposed approach?
Our technology partners, OTS Lab and Digital Roads, have developed two solutions for digital city modelling.
For transport planning of public transport routes and other mobility modes, we use the Replan.City solution based on multi-agent simulation models. We use an activity-based approach to simulate not only simple, single-mode city trips, but also inter- and multi-modal trips using individual mobility means (IMM), new forms of shared-access transport (shared taxis, car sharing, carpooling, etc.).
To build such models and improve the quality of the modelling, we use data from BigData suppliers - for example, on cellular subscriber movements, on contactless transport payment card validation, GPS data on freight movements, etc. Such comprehensive analysis of historical and live data helps us to build an accurate predictive model of a city dweller’s behaviour and subsequently to formulate a mobility management policy. We have already developed transport models for St. Petersburg, Moscow and Greater Moscow, Vladivostok, Krasnoyarsk, and Samara.
Essentially, you create a digital twin of city dwellers’ mobility which then becomes the basis for the development of the public transport system, kicksharing?
True, but this solution alone is not enough to create a fully-fledged digital twin of the city or develop strategic plans for the road infrastructure upgrade. For complex projects, we subsequently engage our partner team, Digital Roads. Their solution, styled Autodiscovery, solves several problems facing city authorities: monitoring the condition of road and transport infrastructure facilities at all life cycle stages, and building a single and up-to-date database of RTI facilities overseen by the agencies responsible. For example, different organisations may possess varying information on the same facilities, which obstructs the development of the RTI, maintenance and repair of existing street and road network elements. All this leads to high infrastructure maintenance costs and inefficient spending of federal and regional budget funds.
How do you create an Autodiscovery-based digital twin?
Mobile units carrying specialised equipment patrol city streets, automatically digitising assigned road infrastructure facilities. The built-in system monitors for changes in current facility conditions and feeds this information into the database online. Following this, a pre-trained neural network recognises facilities, and digitised data (coordinates, dimensions, photos, etc.) is reflected in the facility data sheet and stored in the geographic information system. Thus, the digital twin aggregates up-to-date information on RTI facilities, with coordinates accuracy of up to 10 centimetres. Each mobile unit digitises up to 150 kilometres of road infrastructure every day.
You mentioned traffic enforcement cameras as a road safety improvement measure. What other solutions exist?
The effectiveness of traffic enforcement cameras was confirmed in a recent study undertaken by the Higher School of Economics’ Institute of Transport, according to which accident rates drop on average 2 times, and fatality rates - 3 times (in camera coverage zones). The experience of using traffic enforcement cameras in Greater Moscow, where the number of fatalities in traffic accidents decreases by 15-20% every year, also confirms this trend.
Based on the existing TEC infrastructure, an ecosystem of solutions for a safe and comfortable urban environment has been developed. The platform collects data required for the operation of the traffic safety analysis module and city smart mobility systems, such as traffic density, road congestion rates, length of traffic jams, number of journeys undertaken by a vehicle, plus average speed in different areas. Furthermore, services have been developed to control the colour scheme of taxis and construction waste movements in the Moscow region.
What tasks does the traffic safety analysis module solve?
Traffic police experts analysing accident causes face several problems: a lack of accurate information about the chronology of events and road positions of accident parties, along with information being collected manually and occasionally recorded incorrectly. Besides which, during road reconstruction, the staking that designates the location of items used to determine accident locations, is often misplaced. As a result, the coordinates of accidents that occurred in one and the same place may be spread over distances up to several kilometres apart.
To tackle this, our team has developed both a single database and a data presentation format that rule out discrepancies in the chronology of events (or staking) and enable high-quality accident analysis. Geotags for linking incidents to specific areas in the analysis module facilitate arriving at the correct conclusion by analysing several years’ worth of data. The platform also highlights frequent accident spots on the map, linking them to infrastructure elements and ground facilities, stores and retrieves all the necessary data on traffic accidents and ongoing safety measures, with an assessment of their effectiveness. In this way, regional office experts of the Ministry of Transport, the traffic police and the Centre for Traffic Management receive an all-encompassing tool for collaboration and the prompt decision-making necessary to reduce traffic accident and mortality rates.
You mentioned control of the taxi colour scheme. How does this work?
In late October, we launched the smart Taxi-Control system in Greater Moscow. It monitors vehicles for their compliance with the exterior standard in effect in Greater Moscow, detects the license plate number and matches it to vehicle details entered into the Greater Moscow taxi register.
If there is a match, the neural network will analyse the vehicle’s exterior, checking its colour and the presence of mandatory design elements. The recognition algorithms detect colour scheme elements with high accuracy. If the scheme does not meet the standard, the system sends a notification to the taxi depot requesting them to amend the vehicle’s exterior.
The passenger taxi exterior standard was first introduced in the Moscow region in 2017 - cars should be white with yellow and grey stripes. In addition, they must have a checker belt and an orange roof light.
The neural network was trained on several thousand reference images of taxis, so that it will be able to detect the colour of even a heavily soiled vehicle. Recognition occurs in the camera’s direct view zone - that is, from 20 to 100 metres - depending on visibility and weather conditions. In November 2021 alone, the system processed around ten thousand violations, some of the most common ones being a missing or damaged yellow stripe and a missing roof light.
What are your company's plans for 2022?
We plan to reinforce our presence in the regions where we already operate - that is, Moscow and Greater Moscow, the Nizhny Novgorod Region, St. Petersburg, and the Leningrad Region. Our priorities will be traffic enforcement camera system implementation projects, building on our experience in Greater Moscow, as well as smart mobility systems, digital twinning of cities and public transport upgrades.