The sharing of infrastructure data, both internally across an organisation and externally between organisations, is a daily-basis task in the railway sector. A number of specialized formats have been created for both rail and non-rail focused purposes. The design of each format primarily attempted to address only the specific requirements inherently relating to that given discipline. Since the railway network and its elements are similar in every country, there was no need to reinvent the wheel.

Business objectives

Investing in a standardised railway data format has provided multiple benefits for the sector, including improved data quality, more efficient business performance, streamlined and re-usable development, integrated IT systems, and return on investments.


The RailTopoModel is a logical object model to standardise the representation of railway infrastructure-related data. Together with railML®, which defines the schema for the exchange of data, it has revolutionized the sharing of information in the railway industry.

IRS 30100

Standards in the railway sector have been compiled for more than 100 years in so called UIC leaflets. Today UIC, the largest railway organisation worldwide, calls these regulations International Railway Standards (IRS).
The RailTopoModel 1.0 became such an IRS in spring 2016 and has been IRS 30100 since then.

  • Download RailTopoModel
  • RailTopoModel ↔ railML
  • SNCF's Ariane SOA
  • RailTopoModel and derivatives

Download railTopoModel

RailTopoModel Logo

The different versions of RailTopoModel as well as the final version of the IRS 30100 are available now for the railway sector. You may download it below for free.

For better comprehension, you may inspect the RailTopoModel's wiki too.

Currently we're using the constricted CC licence of railML.org.

RailTopoModel and railML: What's the difference?


RailTopoModel is a generic railway data model that has been designed to support current and future business usages and needs. For example, railway infrastructure managers may set up their asset data base following the basic structure of the RailTopoModel. On the other hand, railML is a common XML-based railway data exchange format. It can be seen as a direct use case of RailTopoModel as it implements the concepts of RTM for data exchange purposes. Together, RailTopoModel and railML 3.x  form the standardized data exchange format proposal.

SNCF's Ariane Service-Oriented Architecture

SNCF Réseau's Ariane Railways Objects Model and Gaïa
SNCF Réseau's Ariane Railways Objects Model and Gaïa

SNCF Réseau has worked since 2011 to develop its global model for railways business objects, Ariane Model, based on the same principles as RailTopoModel. Ariane Model now covers the description of all functional objects and properties of the network, at track and lines levels, including topology, referencing and topography. On 2015, fine routes description and signaling are being added, 2016-2017 will see the enrichment of catenary and power supply network, and finalization of rolling stock (for compatibility).

In parallel, the trains (circulation) and works (engineering-maintenance projects) life cycle objects have been added to Ariane, so it will at the end include all objects related to Infrastructure Management and Operations. Based on Ariane Model, the global repository for Infrastructure, Gaïa, is designed to serve all new IT applications, in a full Object and Service Oriented Architecture. On same principles, an Operational Data Store (ODS) is implemented to collect, and structure on the Ariane common language, all business events processed by legacy systems related to capacity planning, train operations, and works planning. This ODS is progressively the base for development of new applications in those domains.

A major future development taking profit of those bricks (all objects and data based on a common language), will be simulation tools for both design and operations (simulate as it will operate).

RailTopoModel and derivatives

RailTopoModel describes the topology of the railway network in a general way. It is the basis for development of railML 3.x infrastructure scheme. RailTopoModel can also be the base for your infrastructure database. As this database is usually something internal, there is no need for a certification. For certain purposes it may be necessary to extend RailTopoModel and derive modified or enhanced models. You are free to use and extend the model in such a way without further notice to us. However, we will be glad to hear about your approaches and cooperate with you for implementation.

One example for such a RailTopoModel model fork is the RailSystemModel, which has been developed within the EU-funded LinX4Rail project.