The relevance of using computational models for transport planning — both at the scale of industrial enterprises and the national transport network — is discussed by Vladimir Waldin in the ROLLINGSTOCK almanac. Waldin, who is responsible for public transport solutions at Transproekt Ltd., representing the PTV Group holding in the Caspian region, shares his insights.
Published in the almanac “Rolling Stock Market. Kazakhstan” for the TransLogistica Kazakhstan transport and logistics exhibition
Unlike passenger transport, freight transportation by all modes and the entire freight logistics sector are always profit-oriented, or at least aimed at investment payback over time. This is the foundation of the industry’s economics. Therefore, when working with costly transport infrastructure and large fleets, it is crucial to plan many steps ahead to ensure the right choice of geography and engineering investment parameters, as well as the ability to provide the changing market with optimal services—especially in a competitive environment. Perhaps the best example demonstrating the consequences of lacking long-term planning in this area, regardless of the underlying causes, is the fate of rail transport on the American continent. It lost out to road and air transport, as well as to internal competition among parallel companies.
The optimal tool for such planning, which allows “navigation” over any territory, from a region to a country and an entire continent, has been known since the 1960s and has become a classic. It is the mathematical four-step transport demand model, which, as computing power and technology have evolved, has formed the basis of specialised software. This model is the four-step demand model forecasts travel demand by taking into account available transport modes, using four main stages: trip generation (the reason for making a trip), trip distribution (the destination of the trip), mode choice, and trip assignment across the available network routes. Today, the software using this model enables reliable forecasts of transport demand, sufficient for work with both banks and government financial authorities over virtually any planning horizon, depending on the data used in the forecast.
Vladimir Waldin, Transproekt Ltd.
The model is not an oracle and does not provide a definitive answer to what will happen. Multimodal transport modelling employs a scenario-based approach and thus can address any number of “What if…” questions. The list of these “ifs” is virtually endless and may include ongoing territorial development processes, capacities of logistics centres (even those not yet planned), the ongoing shrinking of the Caspian Sea, related climate changes, and handling characteristics. At the national or regional level, models can even assess political risks. The only aspects left unaccounted for are those that were simply not considered or could not be imagined.
On the Eurasian continent, the most widely used solutions for such tasks come from the German company PTV, particularly the PTV Visum. Between 2016 and 2020, the transport components of the European Union’s TriMODE project (Transport Integrated Model of Europe) were implemented using this software, preceded by the SCENES model in 2002, the Eastern Partnership model for the EBRD in 2015, and the national model for the Russian Federation commissioned by the Ministry of Transport in 2020. Germany’s national rail operator Deutsche Bahn, along with its subsidiaries and affiliated companies, has been using PTV software since 1998, running over 1,500 scenario simulations annually. The same software has been used for more than 20 years by national operators of transport and rail infrastructure in Austria, Czechia, Denmark, France, Italy, the Netherlands, Sweden, Switzerland, and the United Kingdom.
The multimodal transport model allows for an accurate reconstruction of the current distribution of freight across transport modes, types, and volumes, and enables the assessment of future demand both in competition and synergy—particularly within diversified holding companies—regarding intermodal container transport. The technology underpinning the model is based on recreating the interaction mechanisms between system users and infrastructure through specialised data sets. This approach not only permits a conditional scoring assessment of the transport situation but also provides a detailed and well-founded forecast. Comparing forecast scenarios helps anticipate and avoid numerous errors by following the principle: “The costliest experiments are conducted on the cheapest (yet reliable and accurate) tool with reversible states and without actual impact on people or the economy”.
A transport model can be broadly represented as a multilayer thematic geographic information system for a specified territory. This system not only includes physically existing objects, areas, points, transport links between them, and their attributes, but also displays possible interaction scenarios among these elements. Using data from this system, the Visum, with its algorithms developed from years of statistical analysis, can simulate the behaviour of transport customers and operators and their interaction with infrastructure, taking into account temporal, cost-related factors, and physical constraints.
Image of the TriMODE transport model (Transport Integrated Model for Europe) in PTV Visum (enlarge). Source: PTV Group presentation
With understanding the entire picture of the current transport configuration and how it is likely to develop in the future, the model user, depending on their capabilities and authority, gains the ability to optimise activities in space and time across a wide range of parameters, up to managing mobility within a region, country, or international project. The concept of mobility management has firmly entered the terminology of urban planners and managers in recent years but fully applies to freight transport of any scale as well.
Macroscopic freight flow management is the prerogative of ministries or intergovernmental bodies, for whom the aforementioned national models were essentially created. At the level of the transport operator, the model allows a “zoom-in” to each specific route and even individual train, enabling work with metrics such as mileage and time, section capacity and throughput, freight turnover, revenue, and operating costs. Meanwhile, the infrastructure operator or investor gains, through the model, a foundation for the technical and economic justification of long-term initiatives, recognised by global audit leaders.
Additionally, depending on the coverage area and level of detail, the model’s functionality provides a foundation for timetable planning (including intermodal schedules), optimising transport schemes considering the location and characteristics of terminals, pricing calculations, and, if necessary, preparing competitive tenders for transportation. The impact of pricing on demand can be analysed within the scope of rail transport or a single operator alone, as well as, given sufficient data (which the model allows to quickly supplement, expand, and maintain across a broad spectrum), in competition with other operators and modes of transport.
The schedule analysis and planning functionality includes the capability to calculate consolidated rolling stock turnover timetables to assess the need for new locomotives and wagons, plan the retirement of older vehicles, and compute annual mileage, wagon-hours, and costs for each train type on every route over an extended period. For example, the Danish national railway company, Danske Statsbaner, conducts such planning under its current contract for the 2025–2035 period. The system supports the consideration of signalling and electrification constraints, and compatibility with external schedule data formats, such as RailML, enables their import and processing to define model sections with different train operating modes for optimisation against various criteria.
The model essentially serves as a digital twin of the transport system, meaning that management decisions made with its aid can be largely automated and become a source of continuously updated data to enhance its performance. This is the goal of another solution, RITM3, developed by the Russian company Simetra. This digital platform manages transport complexes, where the “3” represents “modelling + monitoring + management,” and the acronym stands for Real-Time Integrated in relation to this triad. Major projects implemented using the RITM3 include management systems for two large industrial enterprises with complex external and internal logistics, as well as a unified transport system management platform for a region covering nearly 200,000 km² and a population of over 4 mln people.
Author: Vladimir Waldin, Public Transport Solutions, Transproekt Ltd.
