What we do
1. What we make
SAMRoute produces quantitative risk indicators. Each one describes what surrounds a piece of transport infrastructure, namely the track, the crossing, or the structure, rather than the asset itself. We bring that surrounding context into one analytical view, on the same reference, across the whole portfolio. Owners know their assets. We read the territory around them.
Infrastructure managers' asset-management teams use the output as a portfolio-wide ranking against shared metrics.
Reinsurers' portfolio analysts use it as an independent reading of the asset base they cover.
Indicators reproduce. The same inputs return the same outputs. Every per-asset indicator unfolds back to the foundational metrics, namely counts, distances, and times. It also unfolds back to the statistical estimates we compose from them, such as ratios, and to the assumptions and reference materials that produced each number.
2. The current line of work
Today we describe road–rail level crossings. The work covers every active level crossing on the French national network. Each crossing gets roughly 200 metrics that describe its immediate environment.
The reference materials we build against are public.
CEREMA diagnostic guide — Diagnostic de sécurité routière des passages à niveau — Outils et aide à la réalisation du diagnostic. Covers categories 2 and 3 in dedicated annexes.
EPSF guide INF-GUID-021-V2 — Méthode d'analyse des risques relatifs aux passages à niveau (November 2023). Covers risk analysis.
In late November 2025 we packaged the first complete production-grade run of this work and made it available to SNCF Réseau. Earlier demos in 2025 had set the ground. Presentations followed in December. The Research timeline walks through the year-by-year build-up.
3. What the customer receives
The deliverable is a dataset. For every level crossing in scope, the customer receives a per-crossing record with the full set of metrics. A portfolio-level file aggregates the records across the whole network. The deliverable ships in standard machine-readable formats (CSV, JSON, geo-package). The customer owns the data they receive and can plug it into their own systems.
Two inspection views accompany the dataset for analysts who prefer a browser surface. The portfolio view ranks and filters across the whole network, where the analyst sorts by indicator value, by risk family, by geographical zone, or by combinations. The per-crossing view drills into one crossing and shows the origin–destination pairs that depend on it, with metrics for each pair. The cost ratio quantifies how strongly the surrounding flow depends on the crossing.
4. How it integrates
Since FY2024, the platform pulls in the open and sovereign geospatial layers below. We enrich them with reference taxonomies, run them through the analytical pipeline, and produce the deliverable dataset.
| Source | What it brings | Link |
|---|---|---|
| OpenStreetMap | Road and rail networks, points of interest | openstreetmap.org |
| ESA Copernicus GHSL | Built-up surface layers across Europe | human-settlement.emergency.copernicus.eu |
| ESA Copernicus Worldcover | 10 m global land-cover classification | esa-worldcover.org |
| NASA GPW | Gridded population of the world | sedac.ciesin.columbia.edu |
| IGN RGE Alti | French national elevation model at 1 m and 5 m | geoservices.ign.fr/rgealti |
| Mapillary | Street-level imagery | mapillary.com |
| ONISR | French road-accident reference data | onisr.securite-routiere.gouv.fr |
| SIRENE | French business-establishment register | insee.fr |
| Eurostat | EU statistical units and indicators | ec.europa.eu/eurostat |
| GTFS | Public-transport schedules | gtfs.org |
The Trust & Security page covers EU-only data residency, the sub-processor list, the retention policy, and DPA availability.
The reference dataset starts from the open and sovereign sources alone. Internal customer feeds plug in later, when they add value.
5. Scope today
Today the production scope covers the French national rail network for the level-crossing line of work. We extend to other transport networks one infrastructure manager at a time, at the pace each one can share data.
In the rail-IT stack, the platform sits at the layer that supports decisions and resilience analysis. It complements the methods, audits, and local studies owners already run on their own assets. And it works alongside the operational systems that run the network in the moment, including real-time traffic management, control-command, and ITS.
Every per-asset indicator stays auditable. The reader can unfold it back to the foundational metrics that produced it, namely counts, distances, and times, and to the statistical estimates we compose from them. We document data sources, versions, assumptions, and computation steps for each indicator, available on demand.
6. Reach us
Commercial enquiries, methodology questions, and technical due diligence reach us at contact@samroute.com or +33 2 30 96 66 74.