30/09/2025
How passenger trains can contribute to rail maintenance
The Dutch railway network is heavily used and exposed daily to weather, wear and dynamic loads. Regular maintenance is essential to safeguard safety and reliability. Traditionally, ProRail uses dedicated measurement trains to monitor track condition. But these require dedicated space in the timetable – which is increasingly scarce.
To relieve that pressure, a smart idea emerged: what if we equipped regular passenger trains with measurement technology? These trains already run according to the timetable, so this approach could both reduce planning pressure and increase measurement frequency. But that leads to a fundamental question: how many trains need to be equipped, and which types are best suited for the job?
A complex planning challenge
From the outset, it was clear that the timetable would not be adjusted for measurement purposes. This raised both practical and strategic questions. Do train units typically stick to fixed routes, or are they frequently rotated? And how much of the network is actually ‘covered’ if measurement happens passively, without directing the trains?
The core metric became coverage: what portion of the network is traversed by at least one equipped train within a defined period? But it’s not only about percentage – the reliability of that coverage also matters. This led to a three-part evaluation framework: time horizon, confidence level and coverage.
At the same time: the higher the desired coverage, the greater the investment. But where is the sweet spot? When is good enough truly good enough – within realistic budget constraints?
Data-driven insights based on actual usage
Together with partner Lynxx, CQM helped reduce the complexity to its essence. Using track passage data from ProRail, we analysed how rolling stock actually distributes itself across the network. To keep the problem manageable, we designed a representative ‘minimal’ track map, enabling a simplified but valid reconstruction of the main network flows.
We then ran Monte Carlo simulations: thousands of random combinations of train units were selected, and their combined coverage over a two-week period was calculated. This produced probability distributions per configuration: providing visual and statistical insights into coverage, effectiveness, and reliability. An example of such a distribution is shown in Figure 1.
Figure 1 Estimated distribution of joint network coverage by two VIRM train units over two weeks.
Strategic insight to support future-proof decisions
The analysis revealed how different configuration choices impact overall track coverage. This enables ProRail to make well-informed investment decisions, grounded in both data and practical feasibility.
“The model gives us clear insight into the real-world effect of different scenarios. That allows us to make decisions that are not only technically sound, but also strategically well-founded.”— Bojan Bogojević, Systems Specialist, ProRail
The optimal measurement configuration depends on three key factors:
- Fleet diversity
Different train series follow different usage patterns and together cover more of the network. - Number of equipped trains
More measurement trains increase coverage, but marginal returns diminish with each additional unit. - Design and integration costs
Each train type requires a custom installation design, representing a significant cost factor. Installation itself also adds to the total investment.
These insights make it possible to evaluate different scenarios: what do they deliver, what do they cost, and where does the return make sense?
Because final decisions regarding costs and quality standards were still pending at the time of analysis, we explored a range of scenarios to illustrate possible trade-offs (see Figure 2 and Table 1). Each scenario shows how a specific train type and quantity translates into a certain level of coverage and reliability.
Figure 2 Coverage vs. confidence level for a selection of train configuration scenarios.
Table 1: Key data for selected configuration scenarios.
The added value of collaboration
CQM has over 25 years of experience in solving complex challenges within the rail sector, and knows the dynamics of organisations like ProRail, NS and KeyRail inside and out. In this project, that experience was strengthened by close collaboration with Lynxx and a dedicated team from ProRail.
Merel Groen (Lynxx) contributed deep expertise in rail operations and data infrastructure, while Pepijn Wissing (CQM) brought in his background in statistics, simulation and optimisation. Through frequent on-site alignment, the team was able to share knowledge quickly, validate results, and translate them into applicable insights.
The analysis was well received by both ProRail and NS. A follow-up project is now underway, focusing on technical considerations related to on-board system integration – aiming for a full decision framework for future measurement strategies.
Want to know more?
Curious how CQM uses data, simulation and optimisation to support better decision-making – in your domain too?
Get in touch with Pepijn Wissing or explore our other rail projects at www.cqm.nl
Image by: NS (edited by CQM)
