Publishing of railML 3.2 beta2

by Milan Woelke ( (comments: 0)

As promised during the railML® conference in Gothenburg, Sweden in November 2021, we are happy to inform about the publishing of another beta of the upcoming railML® 3.2. The new version 3.2 beta2 can be downloaded from our website free of charge after login.

It includes updates for the infrastructure, interlocking and rolling stock subschemas compared with the beta1 and introduces the first preview of the new railML® 3 timetable modelling. This version represents another step towards a full railML® 3 version that enables you to describe and exchange data regarding all aspects of modern railway operations. We are continuously working on improving the modelling with support of our stakeholders and partners and are on track to publish the full railML 3.2 version during our next next conference in April 2022.

The following changes have been incorporated into the railML's 3.2 beta 2 version:


railML use case working groups ETCS, ITMS and SCTP have worked on many aspects of the infrastructure domain focusing on clarifying definitions and unambiguous modelling concepts. In particular, working groups contributed to the further development of the elements required for exchange of ETCS related infrastructure data, such as Balise and BaliseGroup locations, ETCS-LevelTransition, SpeedSection and SpeedProfile. Due to the big amount of modifications that all need to pass the proven railML development stages (working group discussion, forum feedback, ticket/issue, modelling, documentation), many approaches are not included in the beta2, but will be part of the beta3 version.

In detail, the following modifications have been done in the railML 3 infrastructure scheme:

  • complete restructuring of Balise and BaliseGroup model including sub-type for Eurobalises and best practice examples
  • complete restructuring of SpeedProfile modelling including best practice examples
  • more detailed modelling of platforms and platformEdges considering operational and geometric aspects
  • introduction of mileageChange as new functional infrastructure element
  • clarification of model for (simple and double) switch crossings in the infrastructure and interlocking domain including best practice examples
  • introducing Areas as flexible location-based elements for various purposes ranging from ETCS to Passenger Information Systems


With the new version the interlocking subschema introduces support for TrainNumberFields based on TvdSections, a feature requested by the community. Support for RBC infrastructure in interlocking was improved, along with the ability to exchange configuration data of involved systems. In general the structure of modeling was improved and naming of elements changed so that misunderstandings of the model can be avoided from the start.

In detail, the following works have been done:

  • abbreviations in element and attribute names eliminated
  • container structure of AssetsForInterlocking allowing more than one set per file
  • RouteRelation for modelling of flank protection
  • TrainNumberFields for indication related to TvdSections
  • ComputerNode data type with IP settings as base type used for SignalBox, RadioBlockCentre and Controller
  • RadioBlockCentre as new main type with general ETCS values
  • OperatorCommands and Indications as additional for each asset/interlocking object
  • EndOfTrack and TrackIL as new assets just referring to infrastructure
  • BasculeBridges and TunnelGateIL as new assets


With version 3.2, the rollingstock subschema that was included in the previous version 3.1 as a stub only, is newly introduced. It supports basic information about vehicles and formations as well as allows description of passenger- and freight facilities.

In detail, the following issues have been addressed:

  • basic information about formations and vehicles
  • available passengerFacilities including information on seats and their classes, services of the coaches, such as air conditioning or restaurant facilities
  • available freightFacilities
  • engine and brake data including use of SegmentTable or ValueTable


As presented during the conference last month work on the completely new timetable modelling has greatly improved. A detailed overview of the current possibilities will be given at an upcoming workshop in January.

The new approach is focused on extendibility as well as minimizing redundancy which was perceived as a major downside of the railML 2.x modelling. Therefore the new timetable model is based on the reuse of train itineraries that can be combined with each other to cover variation without the need for a complete redefinition of a train's path. We still maintain the distinction between the commercial and the operational view of train operations, this approach was even strengthened by separate modellings which allowed us to focus on the relevant data while avoiding ambiguities.

In detail timetable so far includes facilities to exchange data regarding:

  • detailed description of the path between two operational points including applied supplements and their underlying calculation rules
  • details description of the activities at a stop
  • connecting trains
  • dependencies between trains beyond connection
  • mapping of commercial onto operational trains
  • required facilities for a train as well as the available facilities
  • detailed validities as known from railML 2.x


The next step towards the final railML 3.2 release would be the publishing of railML 3.2 beta3 in mid-February 2022. This version will include all features collected in the various working groups. We would very much appreciate it if you could review this new beta2 version of railML and let us know your feedback on what is good and what is missing no later than the 10th of February so that we have time to incorporate your input into the new version of railML. Version beta3 will be followed by the release candidate at the end of March 2022 but will not include features requested after the 10th of February. This will allow us to follow our proven development guidelines to maintain a high level of quality in our work as you have come to expect from us.

Go back