RCM-DX-Specification_EN.md

Specification of the RCM-DX Format

License

TODO: Lizenzhinweise notieren für RCM-DX TODO: Copyright hier vermerken TODO: Lizenzhinweise oder Copyright notieren für HDF5

Contribution

The RCM-DX specification is open source and freely accessible and usable by all with respect to the license. Any person who wants to improve this specification can do so. How exactly this works can be read on the Github website. The repository can be found at RCM-DX repository and the website at RCM-DX specification website.

Change history

RCM-DX version	Document version	Date	Autor	Change
1.0.0	0.1	06.03.2015	Martin Frey (SCS)	Initial version
1.0.0	0.2	20.03.2015	Martin Frey (SCS)	Extensions
1.0.0	0.3	15.04.2015	Patrik Wernli (SCS)	Review
1.0.0	0.4	20.04.2015	Martin Frey (SCS)	Extensions and revised
1.0.0	0.5	03.05.2015	Patrik Wernli (SCS)	Formal Adaptions
1.0.0	0.6	12.05.2015	Martin Frey (SCS)	PDR Feedback: Storing of Booleans, comments allowed on all levels, format independent of video codec, flags (including simulation) on session level. Schemas for exceedances, comments and drawings added
1.0.0	0.7	13.07.2015	Patrik Wernli (SCS)	Finalized for CDR
1.0.0	0.8	02.10.2015	Martin Frey (SCS)	Event model added, reference to specification event schema added.
1.0.0	0.9	30.11.2015	Patrik Wernli (SCS)	Adaptions for Infotrans position model. Version concept removed. Event model updated.
1.0.0	0.10	21.12.2015	Martin Frey (SCS)	Review
1.0.0	0.11	21.12.2015	Patrik Wernli (SCS)	Revised after review
1.0.0	0.12	16.02.2016	Pascal Brem (SCS)	Topology model in configuration.
1.0.0	0.13	17.02.2016	Martin Frey (SCS)	Review topology model
1.0.0	0.14	19.02.2016	Pascal Brem (SCS)	Sections added to file format.
1.0.0	0.15	23.02.2016	Martin Frey (SCS)	Global configuration and settings updated
1.0.0	0.16	11.07.2016	Pascal Brem (SCS)	New Hash code attribute for the topology.
1.0.0	0.17	15.07.2016	Pascal Brem (SCS)	New units and data types for positions
1.0.0	0.18	03.01.2018	Pascal Brem (SCS)	New GTG Track Id in the Topology.
1.0.0	0.19	03.01.2018	Pascal Brem (SCS)	Events are stored on session level.
1.1.0	0.20	09.01.2018	Pascal Brem (SCS)	Events and Sections in a group.
1.1.0	0.21	09.01.2018	Patrik Wernli (SCS)	Added chapter “HDF5 File Format Versions”
1.1.0	0.22	11.04.2018	Patrik Wernli (SCS)	Changed document template to official publishing template
1.1.0	0.23	16.08.2018	Pascal Brem (SCS)	Changes in the channel basis definition.
1.1.0	0.24	16.08.2018	Pascal Brem (SCS)	New attributes on the picture block channel.
1.2.0	0.25	04.09.2018	Pascal Brem (SCS)	New minor version.
1.3.0	0.26	28.11.2018	Pascal Brem (SCS)	New minor version for the topology attributes.
1.4.0	0.27	08.01.2019	Pascal Brem (SCS)	New availability group.
1.4.1	0.28	05.06.2019	Pascal Brem (SCS)	New switchtracks in the DfA
2.0.0	0.29	21.04.2020	Michael Ammann (SBB), Jakob Grilj (SBB)	Adaptation of the structure to new requirements. New major release with version number 2.0, due to major changes in the structure and goal for publication of the specification (open source).
2.0.0	0.30	27.10.2020	Aron Serafini (SCS)	Adjust consistency, measurementMode and processing chapters.
2.0.0	0.31	11.03.2021	Aron Serafini (SCS)	Adjust specification to implementation
2.0.0	0.32	31.08.2021	Aron Serafini (SCS)	Fix mistake: Topology is a child of CONFIGURATION, not Session.
2.0.1	1.0	06.04.2022	Aron Serafini (SCS)	Add attribute MeasurementDirectionDependent to Channel. Use a release version for this document.
2.0.1	1.01	05.06.2023	Mathias Vanden Auweele (Infrabel)	Removed 'group' in all titles and replaced with a reference to HDF5 group. Fix definition of UNRELEASED under the CLEARANCE group. Clarify the definition of "Platform". Changed the limitative platform list into an example platform list.
2.0.1	1.02	26.06.2023	Mathias Vanden Auweele (Infrabel)	Fixed formatting issues with document links. Improved definition of ChannelBasis. Fixed documentation bug with MeasurementDirectionDependent attribute specification

Introduction

Motivation

Railroad companies continuously gather data of their rail, overhead line, telecommunications, civil engineering structure and signaling systems by means of mobile and stationary measuring systems. Data flows from these systems through processing units -- which enrich, evaluate and validate the data --, to systems that display the data to subject matter experts and also to systems that automatically analyse it.

This specification defines the rail condition monitoring data exchange format (RCM-DX format) which is a data format optimised for data in the railroad context, i.e. for data points localised within a railroad network. The RCM-DX format is a file format based on the HDF5 specification and defines a structure of HDF5 groups, datasets, and attributes. The document at hands also describes the content of the elements defined. Although the format is open and can in principle be implemented right away by any railroad company, this specification contains a few non-generic elements and naming conventions that are specific to SBB and Infrabel. The reason for this is that any file that adheres to this specification can be used with the RCM Viewer, an application available soon to the public.

The RCM-DX format is a file format detailing the HDF5 format version 2.0. HDF5 was chosen for several reasons, including that it is an open format. HDF5 is a hierarchical data storage where the data in arranged in a tree structure. The HDF5 format is described on the webpage of the HDF5-group, in particular on the site HDF5 file format specification. The HDF5 group offers tools and libraries for various programming languages and operating systems that allow to read and write HDF5 files.

Hints

RCM-DX structure

RCM-DX defines a structure of HDF5 groups, datasets, and attributes that software solutions that use this format must adhere to.
The extension of the specification is permitted. However, it must be taken into account that such data may no longer be read or processed by existing systems.

Versioning

The RCM-DX data format is subject to changes, these are indicated by the version number in the document, see chapter Root. The version number consists of two numbers, separated by dots and is composed as follows: [Major].[Minor].[Feature]. Example: 1.0.0

Major
Defines the main version and indicates when major changes have been made. These are, for example, that changing the basic structure or renaming groups, datasets or attributes (major, minor attribute as an example), which are mandatory.

Minor
Indicates minor changes, such as changing the name of an attribute that is not mandatory or defining new groups, attributes, or datasets. These changes do not affect anything that cannot be read with an existing RCM-DX read-write library.

Feature
Indicates feature changes.

Diagrams

The following figure shows the color coding used in diagrams:

{width=320px}

Definitions

This document defines technical restrictions as well as content descriptions. This chapter provides an overview of the data types used, types of names, and other important points.

File names

Files following the RCM-DX specification get their own defined file extension, which is rcmdx.

Example of a file name: 20201228_081522_TGMS.rcmdx.

Primitive and extended data types

Primitive data types

Below are the supported data types used in this specification including a short description. Not all data types possible with HDF5 are described. The read and write library specifies which data types can be used for data sets and attributes.

Data type	Description	Examples
boolean	Two values are possible, true or false, 0 or 1	true
byte	One byte or an unlimited number of bytes are possible	-
signed integer	Positive and negative values are allowed. Possible bit depths are: 8, 16, 32 or 64	1256442; -62334
unsigned integer	Only positive values are allowed. Possible bit depths are: 8, 16, 32 or 64	1256442; -62334
float	Positive and negative values are allowed. Possible bit depths are: 32 or 64	12.53; -3212.546
string	Multiple characters of undefined length	"RCM-DX is great!"

Extended data types

Extended data types use the primitive data types, but have a more specialized format or meaning. The following is a list of the extended data types used.

Name	Description	Example
Timestamp	Unique and worldwide defined format of a time in nanoseconds since January 1, 1970 00:00 UTC without leap seconds, defined under wikipedia.org/wiki/Timestamp. Datatype is 64 bit unsigned integer	1553237099000000000
Enum	Enumerations are predefined (and always capitalized) strings defined in this specification. Datatype is string	MAX, MIN, RIGHT, LEFT

(HDF5) Group

If we are talking about a group in this document, we mean the groups in HDF5 format (of the type HDF5 Group). These contain additional groups or datasets.

If a name of a group in this document is written in capital letters (for example TOPOLOGY), it is exactly the same as in the RCM-DX file. If the naming of a group is not fixed, the corresponding chapter describes in more detail how the name is composed.

Groups are described in this specification as follows:

Name	Parent object	Optional
GROUP	RCMDX	no

Name
The name of the group.

Group names whose ending is "_NAME" are wildcard names and are replaced as described in the corresponding paragraph. Example: GROUP_NAME

Parent object
A group can be a subgroup of a group, here the name of this group is mentioned. If the name is written in quotation marks, it can be freely chosen by the creator of the file. Without quotation marks, the name of the group is meant.

Optional
If a group is marked as optional (optional = yes), it can be omitted if it is not needed. If a group is marked as non-optional (Optional = no), then this group must exist if the parent group also exists, otherwise it does not.

(HDF5) Attribute

In the RCM-DX, attributes, groups and datasets can be assigned. The names of the attributes are written in the UpperCamelCase-Notation^1^. Attributes are always of type HDF5 attribute unless otherwise specified.

Attributes are described in this specification as follows:

Name	Data type	Parent object	Optional	Description
StartTime	64 bit integer	SESSION_NAME	no	Start time in miliseconds, for example: 1553237099000000000

Name
The name of the attribute.

Data type
Primitive data type of the attribute, this describes the type of the content in the attribute itself.

Parent object
An attribute is always assigned to a group or a data set, here the name of this group or data set is mentioned.

Optional
If an attribute is marked as optional (Optional = yes), it can be omitted if it is not required. If an attribute is marked as non-optional (Optional = no), this attribute must be present if the group or dataset to which the attribute applies is also present, otherwise it is not.

Description
Description and or examples of the attribute.

^1^Upper-Camel-Case-Notation: The Upper Camel Case Notation defines the way a composite name is written. Further information can be found at the following link: Upper Camel Case

(HDF5) Datasets

A channel as a group always has one data set. A channel defines a type of sensor data that is stored in its data set. Several channels form a data source. Further information on the structure is described further down in this specification.

A data set is always of the HDF5 type HDF5 Dataset.
Below is a list of possible ways in which data can be stored in the RCM-DX:

Type of storage	Description
Array	Data array of arbitrary length
Image	An image taken at a defined time
Video	A video that has been streamed into several individual blocks of defined size, split and saved

The datasets are described in the lowerCamelCase-Notation^2^. Datasets are described in this specification as follows:

Name	Data type	Parent object	Optional	Storage type
timestamp	Timestamp	DATASOURCE_NAME	no	Array

Name
The name of the data set.

Data type
Primitive data type of the content in the data set, thus the data type of the contained data.

Parent object
A data set is always assigned to a group, here the name of this group is mentioned.

Optional
If the dataset is marked as optional (Optional = yes), it can be omitted if it is not needed. If the dataset is marked as non-optional (Optional = no), this dataset must be present if the parent group also has these datasets, otherwise not.

Storage type
One of the storage types described in this chapter.

Descriptions of the data set are added outside the table.

^2^lower-Camel-Case-Notation: The lower camel case notation defines the way a composite name is written. Further information can be found under the following link: www.wikipedia.org/wiki/Camel_case

HDF5 Chunking

The "HDF5 Chunking" is for data within a data set. This means that the data is divided into blocks, which in turn can be processed independently. This also allows faster access to parts of the data. Whether a splitting is allowed and recommended can be seen with each data set, for example: "HDF5 Chunking" is allowed and recommended. The HDF5 chunking is described in more detail on the website of the HDF5 group: www.support.hdfgroup.org/HDF5/doc/H5.user/Chunking.html

Data structures

Within a channel group, one of the following structures can be contained: Array, Limits, Coordinates, Pictures or Videos. These structures are described in more detail in this chapter.

Array

Channels which record individual measured values contain a data set with the name data, this data set is never optional. Single values are stored in this data set as a 1D array, the length of this array (or list) is not limited.

The possible data types are defined in chapter Primitiv and extended data types

Multidimensional measured values are given their own channel group per dimension and thus their own data set called data.

Name	Data type	Parent object	Optional	Storage type
data	A primitive or extended data type	CHANNEL_NAME	no	Array

The following attributes are assigned to this type of data set:

Name	Data type	Parent object	Optional	Description
Unit	string	data set data	no	A physical unit or empty if the data does not correspond to a physical unit

Limits

A channel group can contain zero or one limit groups. Each limit sub-group LIMIT_NAME contains its own timestamp data set and contains a duration data set. The data set limitvalue contains the limit values. When which limit value is valid is defined in the dataset timestamp and how long it is valid is defined in the dataset duration. Values at the same index position of all three data resources belong together.

{width=320px}

The group of limit values is defined as follows:

Name	Parent object	Optional
LIMIT	CHANNEL_NAME	no

The group LIMIT now contains further groups, each with the name of the limit exceeding:

Name	Parent object	Optional
LIMIT_NAME	LIMIT	no

The following attributes are assigned to this group:

Name	Data type	Parent object	Optional	Description
Priority	8 bit integer	LIMIT_NAME	no	Priority of defined limit, lower values are higher priorities.
LimitBound	Enum	LIMIT_NAME	no	Defines the type of limit, possible values are MAX or MIN.

It contains the following datasets:

Name	Data type	Parent object	Optional	Storage type
limitvalue	A primitive or extended data type	LIMIT_NAME	no	Array
timestamp	Timestamp	LIMIT_NAME	no	Array
duration	Timestamp	LIMIT_NAME	no	Array

Coordinates

Measurement data that can be assigned to a coordinate system are given a defined name according to the following pattern: coord.CN.
This type of data storage allows several entries to be recorded per measurement timestamp. Thus there are more entries in these datasets than in the data set timestamp. How many entries per timestamp belong to each other (as a group) is stored in another data set with the name sampleindex. The data set sampleindex is described in chapter Sample index.

Element	Description
coord	Simple character string for identifying data of type Coordinates
.	Separators
C	Additional character for identifying data of type Coordinates
N	Index number beginning with "0", increasing for each additional coordinate datasets

The data set is defined as follows:

Name	Data type	Parent object	Optional	Storage type
coord.CN	A primitive or extended data type	CHANNEL_NAME	no	Single

The following attributes are assigned to this type of data set coord.CN:

Name	Data type	Parent object	Optional	Description
Unit	string	Dataset coord.CN	no	One physical unit or empty if the data does not have any physical unit

Coordinate related measured values

Further measured values can be recorded for each coordinate measuring point. These are stored in individual data sets. In the following the definition of these data sets:

Element	Description
value	Simple character string for identifying data of type Coordinates
.	Separators
V	Additional character for identifying data of type value
N	Index number beginning with "0", increasing for each additional value set

The data set is defined as follows:

Name	Data type	Parent object	Optional	Storage type
value.VN	A primitive or extended data type	CHANNEL_NAME	yes	Single

The following attributes are assigned to this type of data set value.VN:

Name	Data type	Parent object	Optional	Description
Type	string	Dataset value.VN	no	Describes the content and type of the data it contains.

Sample Index

If datasets are created for coordinates, a data set on the same level and with the name sampleindex must be available. The index number of an entry in coord.CN, is entered as the start of the next group. If the data set sampleindex has a value of $21$ at index zero, the first $20$ entries from the data set coord.CN belong together, the next data group start at with index number $21$. The number of entries in sampleindex corresponds to the one in the data set timestamp.

The group sizes can vary among themselves, this can be seen from the index positions in sampleIndex.

Example
The rail cross profile serves as an example here. At one point on the railroad track, several data points of a rail profile are measured and stored. A channel with the name coord.C0 for the X-axis and coord.C1 is created for the Y-axis.

The data set sampleindex now contains the number of entries that belong together.

Below is a picture of a rail cross section measurement with about 2000 measuring points:

Pictures

Images can be saved in compressed or uncompressed form. The format of the images is stored in an attribute so that the image can be read correctly.
Images are stored as binary data blocks, so an image results in a data set. All images are stored in a group called IMG. All images in this group have the same properties that are stored in the attributes.

Name	Parent object	Optional
IMG	DATASOURCE_NAME	no

The group IMG gets the following attributes for the more detailed description of the images contained therein:

Name	Data type	Parent object	Optional	Description
ImageContentType	string	IMG	yes	Image content type, for example JPEG
ResolutionType	Enum	IMG	no	Description in chapter Image resolution types
ResolutionInfoX	32 bit float	IMG	no	Resolution in X direction
ResolutionInfoY	32 bit float	IMG	no	Resolution in Y-direction

Images can have different resolutions in X and Y direction, this must be considered for a correct representation and evaluation of the images.

Image resolution types

The 'ResolutionType' attribute contains information about the values of the 'ResolutionInfoX' and 'ResolutionInfoY' attributes.
ResolutionType' can contain the following values: 'none', 'mmPerPixel' or 'dimension

none
No information available.

mmPerPixel
Defines the pixel size in mm. The values in ResolutionInfoX and ResolutionInfoY give information about the real size of a pixel in millimeters.

dimension
Defines the resolution of the image. The values in ResolutionInfoX and ResolutionInfoY give information about the image size in pixels.

Naming the data set for an image

Name	Data type	Parent object	Optional	Storage type
img.NNNNNNNNN	integer, bit depth depending on color depth	IMG	no	image

The images are named according to the following pattern: img.NNNNNNNNN, hereinafter a description of the individual elements.

Element	Description
img	String for the name of an image
.	Separators.
NNNNNNNNN	Picture number, beginning with 000000000 (nine characters)

Videos

Name	Data type	Parent object	Optional	Storage type
vid.NNNNNNNNN	integer, bit depth depending on color depth	VID	no	image

As with the images, videos can be saved in compressed or uncompressed form. Videos are stored as streams in individual blocks. The blocks are single datasets with a given name.

Name	Parent object	Optional
VID	DATASOURCE_NAME	no

Name of the data set for a video

A video data block is named according to the following pattern: vid.NNNNNNNNN, hereinafter a description of the individual elements.

Element	Description
vid	String for the name of a video
.	Separator
NNNNNNNNN	Video number, starting with 000000000 (nine characters), ascending +1

Time-based data structures

Timestamp

Each entry in a data set of a channel has a reference to an entry in a data set with the name timestamp, which lies within the data source group. In this timestamp data set, there are as many entries as there are entries in a data set of a channel.

The time stamps are always stored in ascending order.

Name	Data type	Parent object	Optional	Storage type
timestamp	Timestamp	DATASOURCE_NAME	no	Array

These time stamps are recorded either by a defined distance travelled or by a frequency, this is described in more detail in the chapter Trigger mode.

Durations

If data is recorded that is valid for a certain period of time, the data set with the name duration is added to the data set timestamp. The timestamp recorded in the timestamp data set specifies the time at which the value was recorded and the duration data set specifies how long this value is valid in nanoseconds. The data set duration is timestamp within a data source group next to the data set.

The differentiation between discrete (data for discrete timestamp) and continuous (data for timestamp with duration) data is done on data source level. The existence or absence of a duration array defines if the data source is "Discrete" (no duration array) or "Continuous" (with duration array).

Example: Assuming there is a data source with a temperature value every second and a calculated average temperature for every minute. Such an average temperature would be stored in a continuous data source, within an array, with a duration of 60 sec.

Name	Data type	Parent object	Optional	Storage type
duration	64 bit integer	DATASOURCE_NAME	no for continuous values, otherwise yes	Array

RCM-DX file format

The RCM-DX consists of a file format of the HDF5-group www.hdfgroup.org/HDF5 in version 2.0. This allows to save the data in a tree structure. This structure, or rather the naming of the groups and datasets, is not specified by the HDF5 group, but by the RCM-DX specified here. The datasets can hold different data, what exactly is contained is specified as metadata.

A change to the structure means a new version and thus a new release of this specification.

To read and write the HDF5 file format, the HDF5 group offers libraries for different languages. These can read and write the structure specified here without problems.

Further information about the structure of the HDF5 file format can be found under the following link: www.portal.hdfgroup.org/display/HDF5/Introduction+to+HDF5

RCM-DX data hierarchy

In the RCM-DX, the individual groups and datasets as well as their names are defined. Below is an overview of the structure specified in this document:

Separate and more detailed specifications have been written for individual structure groups. Several measuring instruments can be installed on one measuring platform. Each of these measuring devices generates new channels of data, which flow into the RCM-DX. Since these channels can be different for each measuring device, the specifications were separated. Another reason for this is the fact that other railway operators use different measuring and inspection equipment.

The individual groups are specified in more detail below in the subcategories.

Root

Root is of type 'Group' in HDF5.

The root group contains all other subgroups. This group defines the RCM-DX and bears its name and thus refers to this specification.

Name	Parent object	Optional
RCMDX	none (this is the root node)	no

{width=400px}

Attributes

The following attributes are assigned to the group RCMDX:

Name	Data type	Parent object	Optional	Description
Major	16 bit integer	RCMDX	no	Major Version of the RCM-DX specification that corresponds to the structure of the created file
Minor	16 bit integer	RCMDX	no	Minor Version of the RCM-DX specification that corresponds to the structure of the created file
Feature	16 bit integer	RCMDX	no	Feature Version of the RCM-DX specification that corresponds to the structure of the created file

File

File is of type 'Group' in HDF5.

The file group contains file specific information, i.e. information specific for the given data set in its current processing state.

Name	Parent object	Optional
FILE	RCMDX	no

{width=300px}

Attributes

The file group contains the following attribute:

Name	Data type	Optional	Description
Element	string	no	Equals to "File". Identifies this node.
StructureVersion	string	no	Version identifier of the platform structure. All underlying systems, datasources and channels can be identified based on this version

Data Processing

Data Processing is of type 'Group' in HDF5.

The data source group DATAPROCESSING contains information about the data in this file, and the processing.

Name	Parent object	Optional
DATAPROCESSING	FILE	yes

Processing log

Processing log is of type 'Group' in HDF5.

The data source group PROCESSINGLOG contains information on data processing. This information is written by systems that make changes to the data. These changes, for example, can be a conversion from millimeters to meters.

Name	Parent object	Optional
PROCESSINGLOG	DATAPROCESSING	yes

Processing log datasets

The group PROCESSINGLOG contains one dataset:

Name	Data type	Parent object	Optional	Storage type
Element	string	no	Equals to "ProcessingLog". Identifies this node.
type	string	PROCESSINGLOG	no	Array
timestamp	64 bit integer	PROCESSINGLOG	no	Array
service	string	PROCESSINGLOG	no	Array
serviceUser	string	PROCESSINGLOG	no	Array
host	string	PROCESSINGLOG	no	Array
executable	string	PROCESSINGLOG	no	Array
user	string	PROCESSINGLOG	yes	Array
message	string	PROCESSINGLOG	yes	Array
All these datasets represent different columns in a table. Their sizes will therefore always be identical.

type
This record contains the type of the data processing step. The following are possible types: CREATION, CONVERSION, MERGE, CONSISTENCY_CHECK, UNIT_CONVERSION, COMMENT, ...

timestamp
Contains the time of the acquisition of the entry.

service
Name of the service that applied this processing step.

serviceUser
(Technical) User that runs the service.

host
Host that the service is run on.

executable
Executable that the service is run from.

user
Optional user-ID of the initiator of this processing step.

message
Optional message of the user.

Clearance

Clearance is of type 'Group' in HDF5.

This group is used by SBB to record information about the data release of all parties who have processed this data. The information is stored in the form of key-value pairs in a data set.

Name	Parent object	Optional
CLEARANCE	DATAPROCESSING	yes

The group CLEARANCE contains one dataset:

Name	Data type	Parent object	Optional	Storage type
Element	string	no	Equals to "Clearance". Identifies this node.
clearance	Enum	CLEARANCE	no	Array
timestamp	64 bit integer	CLEARANCE	no	Array
user	string	CLEARANCE	no	Array
message	string	CLEARANCE	yes	Array
All these datasets represent different columns in a table. Their sizes will therefore always be identical.

clearance
This record contains the enum value of the clearance. The following values are possible:

Name	Description
RELEASED	This file has been released
PENDING	The clearance has not been set yet
UNRELEASED	This file won't be released

timestamp
Contains the time of the acquisition of the entry.

user
User-ID of the initiator of this clearance.

message
Optional message of the user.

Platform

Platform is of type 'Group' in HDF5.

A platform group contains information about a measuring vehicle, mobile or handheld device that collects the data.
The naming of the group is defined according to which platform produced the data. An example list of platforms and associated data is provided in the Platforms example list chapter.

Name	Parent object	Optional
Platform	RCMDX	no

Attributes

The platform group contains the following attributes:

Name	Data type	Optional	Description
Element	string	no	Equals to "Platform". Identifies this node.
VehicleNumber	string	no	Unique number of the vehicle

Platforms example list

Below is a list of the defined unique names of the platforms, their abbreviation and the associated vehicle number.

Platform Name	Abbreviation	Vehicle Number
DFZ00	DFZ	-
DFZ01	gDFZ	-
EM130	EM130	99 88 9162001-0
EM201	EM201	99 88 9163201-5

The Vehicle number is the European Vehicle Number (EVN).

Session

Session is of type 'Group' in HDF5.

The session group contains data that was collected during the same period. A session group contains data from different sources. A RCM-DX file contains exactly one session group.

Naming

The name of a session group is assigned according to the following pattern:

Name	Parent object	Optional
YYYYMMDD_hhmmss.SSS	PLATFORM	no

Example: 20190212_231255.592

The individual elements and their meaning are described below:

Pattern	Content
YYYY	The year in four digit representation
MM	The month in the year (01 for January)
DD	The day in the month
hh	The hour in the day (0-23)
mm	The minute in the hour
ss	The seconds in the minute
SSS	The milliseconds in the seconds
"_" or "."	Characters as separator

For a certain period of time, only one session can exist in a file, this must be ensured by the creator of the file.

Attributes

Name	Data type	Parent object	Optional	Description
Element	string	SESSION_NAME	no	Names the type of the group, this is fix "Session"
StartTime	Timestamp	SESSION_NAME	no	Timestamp in nanoseconds as start time of the session
EndTime	Timestamp	SESSION_NAME	yes	Timestamp in nanoseconds as end time of the session. If the session has not yet been closed, this attribute is missing

Session configuration

Session configuration is of type 'Group' in HDF5.

Configurations can be stored in the datasets of this group. The datasets are designed so that global and network specific configurations can be stored. The configuration can change and have not to be the same in each session.

Name	Parent object	Optional
CONFIGURATION	SESSION_NAME	no

Topology

Topology is of type 'Group' in HDF5.

A topology group contains all information on the route network of the respective railway company.
This chapter has been optimised for SBB and may differ between railway companies. SBB's data processing chain provides for this structure, which is why it is described here.

Name	HDF5 Type	Parent object	Optional
TOPOLOGY	HDF5 Group	CONFIGURATION	no

Attributes

The group TOPOLOGY contains the following attributes:

Name	Data type	Parent object	Optional	Storage type
Element	string	TOPOLOGY	no	Equals to "Topology". Identifies this node.
DfAVersion	string	TOPOLOGY	no	Array
diamond-*	string	TOPOLOGY	no	Array

DfAVersion
Version number of topology, included to check validity.

The DfA (Database of fixed assets) is a SBB construct and reflects the SBB route network. The data comes from a database and is distributed as a file to the SBB measuring vehicles. They can read the information contained therein and also add it to the RCM-DX. This DfA is used for positioning and it is therefore possible to assign the measured data to an object from the route network.

Diamond-Metadata Attributes

List of attributes defined by SBB Diamond. All attributes start with “diamond-“ and have a String value.

Track

Track is of type 'Group' in HDF5.

This group contains information on the tracks of the railway network. The information is stored in separate datasets.

Name	Parent object	Optional
TRACK	TOPOLOGY	no

The following datasets are included in this group, some of which are described in more detail in the subchapters:

Name	Data type	Parent object	Optional	Storage type
Element	string	TRACK	no	Equals to "TopologyTrack". Identifies this node.
direction	Enum	TRACK	no	Array
id	32 bit integer	TRACK	no	Array
gtgid	string	TRACK	no	Array
length	string	TRACK	no	Array
name	string	TRACK	no	Array
nrline	string	TRACK	no	Array
pointfrom	32 bit integer	TRACK	no	Array
pointto	32 bit integer	TRACK	no	Array
switchtype	8 bit signed integer	TRACK	no	Array
tracktype	8 bit signed integer	TRACK	no	Array

direction
The direction of a switch is specified in this data set.
If the track is of the type "switch", a value other than '-' must be selected here. Which number means what is shown in the following table:

Value	Description
-	No crossover
straight	Straight line switch track
left	Left-handed switch
right	switch running to the right

id
ID of the track.

gtgid
Unique GTG ID of a GTG string, this ID is stored as UUID.

length
The length of the track section.

name
Name of the track section.

pointfrom
ID of the starting point of the track section.

pointto
ID of the end point of the track section.

switchtype
If the track is of the type "Switch", a value greater than zero must be selected here. Which number means what is shown in the following table:

Value	Description
0	Anything but a turnout
1	Simple switch
2	Double switch
3	Simple crossovers
4	Double track connection
5	Double crossover

trackType
The number in the tracktype data set defines the type of track that belongs to it. Which number means what is shown in the following table:

Value	Description
0	Station track
1	Track
2	Switch

Line

Line is of type 'Group' in HDF5.

This group contains information about a line in the route network. The information is stored in separate datasets.

Name	Parent object	Optional
LINE	TOPOLOGY	no

The following datasets are included in this group:

Name	Data type	Parent object	Optional	Storage type
Element	string	Line	no	Equals to "TopologyLine". Identifies this node.
id	32 bit signed integer	LINE	no	Array
name	string	LINE	no	Array
firststation	string	LINE	no	Array
laststation	string	LINE	no	Array
fromkilometer	string	LINE	no	Array
tokilometer	string	LINE	no	Array

id
Defines the ID of the line, this is unique.

name
The name of the line.

firststation
The name of the first station of this line.

laststation
The name of the last station of this line.

km
Start kilometre of the line, expressed in kilometres.

tokilometer
Final kilometer of the line, in kilometers.

Switch Track

Switch track is of type 'Group' in HDF5.

This group contains information about switches in the route network. The information is stored in separate datasets.

Name	Parent object	Optional
SWITCHTRACK	TOPOLOGY	no

The following datasets are included in this group:

Name	Data type	Parent object	Optional	Storage type
Element	string	SWITCHTRACK	no	Equals to "TopologySwitchTrack". Identifies this node.
gleisstrangId	32 bit signed integer	SWITCHTRACK	no	Array
gleisstrangBez	string	SWITCHTRACK	no	Array
weicheId	32 bit signed integer	SWITCHTRACK	no	Array
ablenkRichtung	string	SWITCHTRACK	no	Array
ablenkung	string	SWITCHTRACK	no	Array
betriebspunkt	string	SWITCHTRACK	no	Array
herzStueck	string	SWITCHTRACK	no	Array
minRadius	32 bit signed integer	SWITCHTRACK	no	Array
nr	32 bit signed integer	SWITCHTRACK	no	Array
zusNr	string	SWITCHTRACK	no	Array
schienenProfil	string	SWITCHTRACK	no	Array
status	string	SWITCHTRACK	no	Array
schwellenArt	string	SWITCHTRACK	no	Array
typenPlanNr	32 bit signed integer	SWITCHTRACK	no	Array
typNrZusatz	string	SWITCHTRACK	no	Array
weichenArt	string	SWITCHTRACK	no	Array
weichenTyp	string	SWITCHTRACK	no	Array
weichenForm	string	SWITCHTRACK	no	Array
weichenZunge	string	SWITCHTRACK	no	Array

gleisstrangId
A reference to the GTG-ID.

gleisstrangBez
Contains a description of the track section.

weicheId
Contains the ID's of the switches as a reference.

Track Object

Track object is of type 'Group' in HDF5.

This group contains information about objects in the route network, for example a balise. The information is stored in separate datasets.

Name	Parent object	Optional
TRACKOBJECT	TOPOLOGY	no

The following datasets are included in this group:

Name	Data type	Parent object	Optional	Storage type
Element	string	TRACKOBJECT	no	Equals to "TopologyTrackObject". Identifies this node.
trackid	32 bit signed integer	TRACKOBJECT	no	Array
type	32 bit signed integer	TRACKOBJECT	no	Array
positionstart	32 bit signed integer	TRACKOBJECT	no	Array
positionend	32 bit signed integer	TRACKOBJECT	no	Array
extrainfo	string	TRACKOBJECT	no	Array

trackid
Contains the ID of the track to which the track is connected.

type
Type of the object.

positionstart
Start position of the object in meters.

positionend
End position of the object in meters.

Extrainfo
Additional information about the object, for example, the ID of a balise.

Track Point

Track point is of type 'Group' in HDF5.

This group contains information about defined points on the route network. The information is stored in separate datasets.

Name	Parent object	Optional
TRACKPOINT	TOPOLOGY	no

The following datasets are included in this group:

Name	Data type	Parent object	Optional	Storage type
Element	string	TRACKPOINT	no	Equals to "TopologyTrackPoint". Identifies this node.
trackid	32 bit signed integer	TRACKPOINT	no	Array
lineid	32 bit signed integer	TRACKPOINT	no	Array
x	32 bit float	TRACKPOINT	no	Array
y	32 bit float	TRACKPOINT	no	Array
z	32 bit float	TRACKPOINT	no	Array
radius	32 bit float	TRACKPOINT	no	Array
kilometer	32 bit float	TRACKPOINT	no	Array
position	32 bit float	TRACKPOINT	no	Array
cant	32 bit float	TRACKPOINT	no	Array
steigung	32 bit float	TRACKPOINT	no	Array

trackid
Reference to the ID of the track section.

lineid
Reference to the ID of the line.

x
X-Coordinate of the point.

y
Y-Coordinate of the point.

z
Z-Coordinate of the point.

radius
The radius of a point, given in meters.

kilometer
Contains the line kilometre of the point in the route network, expressed in kilometres.

position
Position of the point, in meters.

cant
The inclination at this point, expressed in millimetres.

steigung
Gradient at this point, expressed in parts per thousand.

Property

Property is of type 'Group' in HDF5.

This group contains information about properties of the topology itself. The information is stored in separate datasets.

Name	Parent object	Optional
PROPERTY	TOPOLOGY	no

The following datasets are included in this group:

Name	Data type	Parent object	Optional	Description
Element	string	PROPERTY	no	Equals to "TopologyProperty". Identifies this node.
id	32 bit signed integer	PROPERTY	no	Array
name	string	PROPERTY	no	Array
description.ge	string	PROPERTY	no	Array
description.fr	string	PROPERTY	no	Array
description.it	string	PROPERTY	no	Array
description.en	string	PROPERTY	yes	Array

id
Unique ID of the characteristic.

name
Name of the characteristic.

description.ge
Description of the characteristic in the language German.

description.fr
Description of the characteristic in French language.

description.it
Description of the characteristic in Italian language.

description.en
Description of the feature in English language.

Setting

Setting is of type 'Group' in HDF5.

Within this group there are further groups whose names identify the type of setting.
In the following SETTING_NAME is used as placeholder of the actual name of a setting.

Name	Parent object	Optional
SETTING_NAME	CONFIGURATION	yes

This group contains two datasets:

Name	Data type	Parent object	Optional	Storage type
Element	string	SETTING_NAME	no	Equals to "Settings". Identifies this node.
setting	string	SETTING_NAME	no	Array
timestamp	64 bit integer	SETTING_NAME	no	Array

setting
Contains the actual configuration.

timestamp
Contains the time from when this configuration is valid and was used.

The following attributes are contained in this group:

Name	Data type	Parent object	Optional	Description
DataType	string	setting	no	Defines the datatype of the configuration within the data set setting. Data type specified as MIME^3^ type, for example Content-Type: <text/strings>

Sections

Section is of type 'Group' in HDF5.

The group SECTIONS, contains information about a session.

Name	Parent object	Optional
SECTIONS	SESSION_NAME	no

Data fields

The following data fields are contained in the group "SECTIONS":

Name	Data type	Parent object	Optional	Storage type
Element	string	SECTIONS	no	Equals to "Sections". Identifies this node.
direction	Enum	SECTIONS	no	Array
startTimestamp	Timestamp	SECTIONS	no	Array
endTimestamp	Timestamp	SECTIONS	no	Array
firstTrackOffset	64 bit float	SECTIONS	no	Array
lastTrackOffset	64 bit float	SECTIONS	no	Array
trackInfoOffset	64 bit float	SECTIONS	no	Array
trackId	32 bit signed integer	SECTIONS	no	Array
trackStartTimestamp	Timestamp	SECTIONS	no	Array
trackEndTimestamp	Timestamp	SECTIONS	no	Array
trackKilometrage	Enum	SECTIONS	no	Array
trackStartCoveredDistance	64 bit float	SECTIONS	no	Array
trackEndCoveredDistance	64 bit float	SECTIONS	no	Array

direction
Indicates the orientation of travel of the measuring vehicle in the given section. This array contains only as many entries as there are sections.

Possible values of this enumeration are:

Value	Orientation of travel
FORWARD	forward
BACKWARD	backward

startTimestamp
Start time of the section as time stamp since 1.1.1970 at 00:00 UTC.

endTimestamp
End time of the section as time stamp since 1.1.1970 at 00:00 UTC. endTimestamp must be greater than startTimestamp.

firstTrackOffset
Indicates the distance in meters between the start of the track and the position at the beginning of the section. This array contains only as many entries as there are sections.

lastTrackOffset
Indicates the distance in meters between the position at the end of the measurement and the end of the rail. This array contains only as many entries as there are sections.

trackInfoOffset
This data set lists how many entries in the datasets of the "Track list group" belong to section. One entry is created per section in a session and the number of entries is defined. The number of tracks belonging to each section can be determined by calculating the specified offset value at the $x$ position minus the offset value at the $x-1$ position.

trackId
ID of the tracks that are part of the section.

trackStartTimestamp
Time since beginning of the session at which the measurement on a track started.

trackEndTimestamp
Time since the beginning of the session at which the measurement on a track ended. endTimestamp must be greater than startTimestamp.

trackKilometrage
Orientation of the track with respect to the section’s driving direction.

The following values are allowed:

Value	Meaning
ASCENDING	The track kilometrage increases with the driving direction
DESCENDING	The track kilometrage decreases with the driving direction

trackStartCoveredDistance
Start covered distance of the track in the section.

trackEndCoveredDistance
End covered distance of the track in the section.

Measuring System

Measurement system is of type 'Group' in HDF5.

Each measuring system has its own data sources, which have their own names, as well as their own channels, which in turn have their own names. Common features are described in this specification, everything else is defined in a separate specification. Since this part differs greatly among railway companies and with measuring equipment, a rigid specification has been dispensed with, but a certain framework is still given.

A group is created for each system that collects data. The name of the group is unique for each system. The composition of this name is not predefined.

Name	Parent object	Optional
MEASURINGSYSTEM_NAME	SESSION	no

{width=400px}

Attributes

The following attributes are contained in the group of the measuring system:

Name	Data type	Parent object	Optional	Description
Element	string	MEASURINGSYSTEM_NAME	no	Indicates the type of the group, this is fixed System

Datasource

Datasource is of type 'Group' in HDF5.

A data source group can contain several channels and thus several data sources. This group combines these channels. The naming can be freely selected, but must be unique.

Name	Parent object	Optional
DATASOURCE_NAME	MEASURINGSYSTEM_NAME	no

A timestamp is available for each individual measuring point within a data source group. There are two types of data acquisition for a data source group. One is always after a defined distance (e.g. every 250 millimeters) and the other is the recording of measurement data at a certain frequency (e.g. 4000 Hz). The way the measurement data was recorded is shown in two attributes for each channel group. For a description see Trigger mode.

{width=320px}

Attributes

The following attribute is assigned to the group:

Name	Data type	Parent object	Optional	Description
Element	string	DATASOURCE_NAME	no	Contains the type of the group, this is fix Datasource

Example

In our example the name of the data source group is assigned, which should contain our environmental measurement data, which we call ENVIRONMENT.

Timestamp data set

Each data source group contains a data set called timestamp. It contains all timestamps at which a measurement was recorded. The size of this list of timestamps is the same as the size of the datasets per channel.

A more detailed description can be found in the chapter Timestamp Array!

Channel

Channel is of type 'Group' in HDF5.

A channel group contains metadata for the actual measurement data available in the channel. The naming can be freely selected, but must be unique within the data source group.

Name	Parent object	Optional
CHANNEL_NAME	DATASOURCE_NAME	no

The following attributes are contained in this group:

Name	Data type	Parent object	Optional	Description
TriggerMode	Enum	CHANNEL_NAME	no	See chapter Trigger Mode
CommonTriggerDistance	64 bit float	CHANNEL_NAME	no	See below
CommonTriggerFrequency	64 bit float	CHANNEL_NAME	no	See below
ChannelBasis	Enum	CHANNEL_NAME	no	See below
MeasurementDirectionDependent	boolean	CHANNEL_NAME	yes	See below
MeasurementType	Enum	CHANNEL_NAME	no	See below
Neighbor	string	CHANNEL_NAME	no	See below
MeasurementUncertaintyType	Enum	CHANNEL_NAME	yes	See below
LowerMeasurementUncertainty	32 bit float	CHANNEL_NAME	yes	See below
UpperMeasurementUncertainty	32 bit float	CHANNEL_NAME	yes	See below
PositionOffset	32 bit signed float	CHANNEL_NAME	no	See below
Element	string	CHANNEL_NAME	no	Contains the type of the group, this is fix Channel

CommonTriggerDistance
Data is collected every value meters.

In a channel group there is always only one common trigger value! The other trigger value is set to "0.0"

Both trigger values have the value "0.0" if TriggerMode contains the value EVENTS.

CommonTriggerFrequency
Data is collected with a frequency of value Hertz.

In a channel group there is always only one common trigger value! The other trigger value is set to "0.0"

Both trigger values have the value "0.0" if TriggerMode contains the value EVENTS.

ChannelBasis
Defines the applicable side of the data signal in relation to the vehicle or moving direction. A sensor can be installed on one side of the platform. Depending on the direction and the orientation of the vehicle, this can be 'left' or 'right' in relation to the movement but is always fixed in relation to the vehicle frame axes.

Possible values are:

Value	Description
TOTAL	The channel has no dependency on direction of motion or vehicle orientation
SENSOR_LEFT	Deprecated – same as SENSOR_VERTICAL_LEFT
SENSOR_RIGHT	Deprecated – same as SENSOR_VERTICAL_RIGHT
RAIL_LEFT	Deprecated – same as MOVE_DIRECTION_VERTICAL_LEFT
RAIL_RIGHT	Deprecated – same as MOVE_DIRECTION_VERTICAL_RIGHT
SENSOR_VERTICAL_LEFT	Signal reflects the left hand side of the vehicle irrespective of it's orientation or motion
SENSOR_VERTICAL_RIGHT	The channel reflects the right-hand side of the vehicle irrespective of its orientation or motion
MOVE_DIRECTION_VERTICAL_LEFT	The channel reflects the left-hand side with respect to direction of motion, irrespective of its orientation
MOVE_DIRECTION_VERTICAL_RIGHT	The channel reflects the right-hand side with respect to direction of motion, irrespective of its orientation
SENSOR_VERTICAL_TOTAL	The channel reflects the center of the vehicle irrespective of its orientation or motion
MOVE_DIRECTION_VERTICAL_TOTAL	The channel reflects the center with respect to direction of motion, irrespective of its orientation

MeasurementDirectionDependent
Flag that implies, if set to true, that measurement data of this channel is dependent on direction and should be inverted if viewed with a section of different direction.

If the flag is set to false, not set or only data of a single direction is viewed, then no inversion is needed when viewing data of this channel.

MeasurementType
Defines how a value was created. This can be measured, calculated or taken from a previously defined data source that was read from there and inserted into the file.

The following values are possible:

Value	Description
MEASURED_VALUE	Measured values
REFERENCE_VALUE	A setpoint of a third source

An example for reference values are defined target values which flow in from another source (file, database etc.) and are to be used for comparisons.

Neighbor
Refers to the name of an adjacent channel. This can be the right rail, for example, when measuring the track temperature of the left rail. Thus the attribute neighbor of the channel "TEMP_RAIL_L" would contain the name "TEMP_RAIL_R" and vice versa.

MeasurementUncertaintyType
Defines the type in which the measurement uncertainty is set, namely absolute (in which case it has the same unit as the channel itself), or relative (in which case it has no unit). The following values are thus allowed:

Value	Description
ABSOLUTE	Absolute values
RELATIVE	Relative values

This attribute is optional. If it is not set, no information is known about the measurement uncertainty.

LowerMeasurementUncertainty
This attribute contains the measurement accuracy of the channel according to the specifications of the measurement system. This is the lower bound for the measurement uncertainty range. This attribute is optional. If it is not set, no information is known about the measurement uncertainty.

UpperMeasurementUncertainty
This attribute contains the measurement accuracy of the channel according to the specifications of the measurement system. This is the upper bound for the measurement uncertainty range. This attribute is optional. If it is not set, no information is known about the measurement uncertainty.

PositionOffset
Describes the distance between a defined zero point (position) on the measuring vehicle and a the measuring sensor. This specification is used to convert the exact time at which the measurement was taken to a defined zero point. The value can be positive or negative and has the unit millimeter.

TriggerMode

This attribute defines how the data was recorded.

Possible values are:

Value	Description	Unit
TIME	Time-based measurement data recording	Nanoseconds
DISTANCE	Distance-based measurement data acquisition	Millimeter
EVENT	EVENT based recording	none

Data object

Each channel group receives a data set with the actual measurement data:

Name	HDF5 Type	Optional
data	HDF5 Dataset	no

There are as many measurement data entries as there are timestamps in the data set timestamp which is included in the channel group.

The data set needs more information, this is given as attributes:

Name	Data type	Parent object	Optional	Storage type
Unit	string	CHANNEL_NAME	no	Array

Unit:
The physical unit of the measurement data, such as "millimeter". If no physical unit can be assigned to the data, this attribute remains empty.

The data set and the possible data that can be stored are described in more detail in the chapter Dataset.

Measurement mode

Measurement mode is of type 'Group' in HDF5.

The following group contains important information about the measurement mode of the system.

Name	Parent object	Optional
MEASUREMENTMODE	MEASURINGSYSTEM_NAME	yes

The group MEASUREMENTMODE contains one dataset:

Name	Data type	Parent object	Optional	Storage type
Element	string	MEASUREMENTMODE	no	Equals to "MeasurementMode". Identifies this node.
mode	Enum	MEASUREMENTMODE	no	Array
timestampfrom	64 bit integer	MEASUREMENTMODE	no	Array
timestampto	64 bit integer	MEASUREMENTMODE	no	Array
info	string	MEASUREMENTMODE	yes	Array
All these datasets represent different columns in a table. Their sizes will therefore always be identical.

Mode (enum)

There are three different measurement modes, which are explained individually below.

Name	Description
PRODUCTIVE	Productive data that will be further used.
TEST	Test data recorded during a diagnostic run with the aim of checking and testing the measuring equipment.
SIMULATION	Simulated values that the measuring systems produce themselves and are no longer used.

Logging

Logging is of type 'Group' in HDF5.

The logging group contains information about the status of the measuring systems. The data is divided into two subgroups, AVAILABILITY and MESSAGES. These are described in separate chapters.

Name	Parent object	Optional
LOGGING	MEASURINGSYSTEM_NAME	yes

{width=320px}

Availability

Availability is of type 'Group' in HDF5.

In this group, failures and interruptions of measurement systems are recorded in a defined structure, each as its own data set.

Name	Parent object	Optional
AVAILABILITY	LOGGING	yes

The following datasets are included in this group:

Name	Data type	Parent object	Optional	Storage type
Element	string	AVAILABILITY	no	Equals to "Availability". Identifies this node.
availability	Enum	AVAILABILITY	no	Array
systemreference	string	AVAILABILITY	no	Array
datasourcereference	string	AVAILABILITY	yes	Array
channelreference	string	AVAILABILITY	yes	Array
timestampfrom	Timestamp	AVAILABILITY	no	Array
timestampto	Timestamp	AVAILABILITY	no	Array
message	string	AVAILABILITY	no	Array

availability
See chapter Availability

systemreference
A reference to the measurement system.

datasourcereference
A reference to a data source.

channelreference
A reference to a channel.

message
This data set contains one message per entry about a failure of a measuring instrument.

timestampfrom
Start timestamp of the availability entry.

timestampto
End timestamp of the availability entry.

Availability

Defines the severity of the failure or interruption of a measurement system. The following values are possible:

level	description
MONITORED	The measurement system or sensor was monitored during this period. All other
FAILURE	The measuring system or sensor has failed and has not recorded any measured values during the session.
WARNING	The measuring system or the sensor may have had a failure and the measured values may not be correct.

Consistency

Consistency is of type 'Group' in HDF5.

The message about the consistency of the data is triggered by a system that checks all data according to certain criteria. For example, this could be a check for black images in a video. If all frames in the video are black, something is wrong and the video is unusable.

Name	Parent object	Optional
CONSISTENCY	LOGGING	yes

The group CONSISTENCY contains one dataset:

Name	Data type	Parent object	Optional	Storage type
Element	string	CONSISTENCY	no	Equals to "Consistency". Identifies this node.
channelid	string	CONSISTENCY	no	Array
timestampfrom	64 bit integer	CONSISTENCY	no	Array
timestampto	64 bit integer	CONSISTENCY	no	Array
consistencytype	Enum	CONSISTENCY	no	Array
consistencyinfo	string	CONSISTENCY	yes	Array
All these datasets represent different columns in a table. Their sizes will therefore always be identical.

channelid
Channel reference in which the finding was detected.

timestampfrom
Start timestamp of the consistency.

timestampto
End timestamp of the consistency. Note that consistencies can never overlap and will always be continuous (without holes) within sections.

consistencytype
The consistency type can have the following values:

Name	Description
CONSISTENT	The referenced data was checked and is consistent
INCONSISTENT	The referenced data was checked and is inconsistent
NO_DATA	Consistency has been checked, but no data was found

consistencyinfo
Contains additional information about this consistency (E.g. which rule decided the consistency type).

Message

Message is of type 'Group' in HDF5.

This group contains messages, generated from a measurement system or a person, structured in data set.

Name	Parent object	Optional
MESSAGES	LOGGING	yes

The following datasets are included in this group:

Name	Data type	Parent object	Optional	Storage type
Element	string	MESSAGES	no	Equals to "Messages". Identifies this node.
message	string	MESSAGES	no	Array
level	Enum	MESSAGES	no	Array
systemreference	string	MESSAGES	no	Array
datasourcereference	string	MESSAGES	no	Array
channelreference	string	MESSAGES	no	Array
timestamp	Timestamp	MESSAGES	no	Array

message
This data set contains one message per entry about a failure of a measuring instrument.

systemreference
A reference to the measurement system.

datasourcereference
A reference to a data source.

channelreference
A reference to a channel.

level
See chapter Message level

Message level

Defines the importance of a message. Following values are possible:

level	description
INFO	The message is only informative.
WARNING	The message indicates a warning.
ERROR	The message indicates an error.

Position

Position is of type 'Group' in HDF5.

This group contains general information on the position. It is like a measuring system (see Measuring System) with following differences:

The channel with the position data has multiple (nine) datasets. This is because a position record always consists of these nine values and therefore this dataset is hardcoded, instead of configurable as for all other measurement system.

The channel and datasets have no attributes (e.g., physical unit) because this information is part of the file format and not configurable.

Name	Parent object	Optional
POSITION	SESSION	no

Position datasource

See chapter Datasource

Name	Parent object	Optional
POSITION.SOURCE	POSITION	no

Position channel

Name	Parent object	Optional
POSITION.SOURCE.DATA	POSITION.SOURCE	no

Position channel datasets

Name	Data type	Parent object	Optional	Storage type
covereddistance	64 bit float	POSITION.SOURCE.DATA	no	Array
direction	Enum	POSITION.SOURCE.DATA	no	Array
kilometrage	Enum	POSITION.SOURCE.DATA	no	Array
track_id	32 bit integer	POSITION.SOURCE.DATA	no	Array
line_id	32 bit integer	POSITION.SOURCE.DATA	no	Array
trackoffset	64 bit float	POSITION.SOURCE.DATA	no	Array
linekilometer	64 bit float	POSITION.SOURCE.DATA	no	Array
positionaccuracy	8 bit integer	POSITION.SOURCE.DATA	no	Array
positionquality	8 bit integer	POSITION.SOURCE.DATA	no	Array

covereddistance
The distance (in [m]) the vehicle is driven at the time of recording within the current session.

track_id
The ID of the track on which the vehicle is located at the time of recording.

line_id
The ID of the line on which the vehicle is located at the time of recording.

trackoffset
The offset (in [m]) on the track where the vehicle is located at the time of recording.

linekilometer
The line kilometer (in [km]) on the line where the vehicle is located at the time of recording.

positionquality
Marker for the quality of the position determination ranging from 0 meaning “very good” to 15 for “very poor”.

positionaccuracy
The position accuracy (in [m]) of the current position.

Coach Direction

The data set direction contains the coach direction of the vehicle. This information influences the position of the measuring systems.

This data set can contain the following values:

Value	Description
FORWARD	Vehicle moving forward
BACKWARD	Vehicle moving backward

Kilometrage

The data set kilometrage contains the alignment of the track. This information serves the correct evaluation of the kilometer data of the line, see data set trackoffset.

This data set can contain the following values:

The following values are allowed:

Value	Meaning
INCREASING	Increasing track kilometrage on the track
DECREASING	Decreasing track kilometrage on the track

EVENTS

Events is of type 'Group' in HDF5.

The EVENTS group is used to store events that occurred during the recording of data. Events are bound to a channel, system or session and have a link to it. Systems can, for example, trigger an EVENT when a limit value is exceeded. Events are always time-bound which means an EVENT contains the exact time of occurrence and the duration of the EVENT. The duration can also be zero, so the EVENT occurred exactly at the specified time.

Name	HDF5 Type	Parent object	Optional
EVENTS	HDF5 Group	SESSION_NAME	yes

Within the group there are the following data fields:

Name	Data type	Parent object	Optional	Storage type
Element	string	`EVENTS``	no	Equals to "Events". Identifies this node.
systemref	string	EVENTS	no	Array
channelref	string	EVENTS	yes	Array
eventtype	string	EVENTS	no	Array
data	string	EVENTS	no	Array
duration	64 bit signed integer	EVENTS	no	Array
timestamp	Timestamp	EVENTS	no	Array

Each of these datasets contains a list with information about an entry, at a certain time. Each data set is described in more detail in the following subchapters.

systemref Contains a list of entries containing the name of the system that triggered the EVENT.

Name	Data type	Parent Object	Optional	Storage Type
systemref	string	EVENTS	no	Array

channelref Contains a list of entries that refers to a channel to which the EVENT applies.

Name	Data type	Parent Object	Optional	Storage Type
channelref	string	EVENTS	yes	Array

data This data set contains the actual information about an EVENT, this in the XML notation which is described in more detail in each chapter of the EVENT types (Defect EVENT type, Detected object EVENT type, Limit EVENT type).
A type can be stored for each EVENT. These are explained in more detail below.

Name	Data type	Parent object	Optional	Storage type
data	string	EVENTS	no	Array

duration Defines for each EVENT the duration of the EVENT itself. This value can also be zero.

Name	Data type	Parent object	Optional	Storage type
duration	64 bit integer	EVENTS	no	Array

type Contains the type of an EVENT.

Name	Data type	Parent object	Optional	Storage type
type	string	EVENTS	no	Array

In the list "type" the type of the recorded EVENT is shown. The different types contain different information which is shown in the following subchapters. There are corresponding XML schemas for all types that define the technical specifications.

Defect EVENT type

A defect can be, for example, an image of a rail showing a damage of the surface. This defect is recorded by a system. However, it may happen that this error is not one (incorrectly detected), this information can be specified afterwards (attribute "PossibleValidationResults"). Defects are always channel bound and recorded or evaluated by a system. In the following, the elements and attributes that occur in a Defect as XML are described in more detail.

The XML Schema can be found in the chapter EventsDefect.

XML elements

Not all of these elements must be present, details can be taken from the XML Schema.

Name	Description	Parent object
Defect	XML Root Element	none
PossibleDefectNames	Name of a possible error	Defect
PossibleClassifications	Classification of a possible defect	Defect
PossibleValidationResults	Possible confirmations of the defect	Defect

XML attributes

Below are the attributes of the root element "Defect":

Name	Description	Parent object
Classification	Classification of the error	Defect
DefectName	Name of the error	Defect
Details	Further information or more detailed descriptions of the error	Defect
Parameter1Name	Name of the parameter 1	Defect
Parameter1Value	Value of parameter 1	Defect
Parameter2Name	Name of the parameter 2	Defect
Parameter2Value	Value of Parameter 2	Defect
Parameter3Name	Name of the parameter 3	Defect
Parameter3Value	Value of parameter 3	Defect
ID	Unique number for identification of the error	Defect

Detected object EVENT type

These events indicate an object found during a diagrose ride. These can be, for example, detected balises or tunnels. What exactly counts as a found object is not defined in this specification, only the information for a recorded EVENT.

The XML Schema can be found in the chapter EventsGeneric.

XML elements

Not all of these elements must be present, details can be taken from the XML schema.

Name	Description	Parent object
DetectedObject	Root Element	none
object	Element with information about the found object in the element itself or in the attributes	DetectedObject
Reference	Reference to a list of known and uniquely assignable objects of the railway company	DetectedObject
ObjectAttribute	Further information about the object, the information is contained in the attributes	DetectedObject

XML attributes

Name	Description	Parent object
Unique ID of the EVENT	DetectedObject
Type	Type of object found	object
Description	further description or information about/from object	object
ObjectConsistency	Reference to the correctness of the specified data	object
ReferenceSystem	Reference to the name of the system from which the data originates	Reference
Key	Information about the data contained in the "ObjectAttribute" element	ObjectAttribute

Limit violation EVENT type

Limit value exceedances of measured values of a channel can also be recorded as events.
The XML schema can be found in chapter Events Generic.

XML elements

Name	Description	Parent object
LimitViolation	Root Element	none

XML attributes

Name	Description	Parent object
TimestampMaxViolation	Time at which limit value was exceeded	LimitViolation
ViolatedLimit	Name of the defined limit	LimitViolation
ID	Unique ID of the EVENTS	LimitViolation

Comment

Comments recorded during a diagnostic drive by the user. The content is not specified, only the XML structure. The XML schema can be found in chapter Events Comment.

XML elements

Name	Description	Parent object
Comment	Root element and message, recorded by the user	none

XML attributes

Name	Description	Parent object
Username	Name of the user who recorded the message	Comment
ID	Unique ID of this message	Comment

Corrupt

Messages of the type "damaged" or "unusable" do not receive a content specification, only the XML structure is predefined and described here. The XML schema can be found in chapter Events Generic.

XML elements

Name	Description	Parent object
Corrupt	Root element and message, recorded by the user	none

XML attributes

Name	Description	Parent object
Username	Name of the user who recorded the message	Corrupt
ID	Unique ID of this message	Corrupt

Measurement system configuration

Measurement system configuration is of type 'Group' in HDF5.

The structure of this group is the same as for the configuration group below the group SESSION_NAME: Session configuration, but the parent group is *MEASURINGSYSTEM_NAME*: Measuring system. The measuring system configuration group also does not contain the Topology.

{width=230px}

XML Schema Definitions

Events Comment

<?xml version="1.0" encoding="UTF-8"?>
<xs:schema xmlns:xs="http://www.w3.org/2001/XMLSchema"
  xmlns:tns="http://www.sbb.ch/RCMDX/Events/Comment"
  targetNamespace="http://www.sbb.ch/RCMDX/Events/Comment"
  elementFormDefault="qualified">

  <xs:include schemaLocation="../RcmDxDataTypes.xsd" />

  <xs:element name="Comment">
    <xs:complexType>
      <xs:simpleContent>
        <xs:extension base="xs:string">
          <xs:attribute name="Username" type="xs:string" use="required" />
          <xs:attribute name="ID" type="tns:UUID" use="required" />
        </xs:extension>
      </xs:simpleContent>
    </xs:complexType>
  </xs:element>

</xs:schema>

Events Defect

<?xml version="1.0" encoding="UTF-8"?>
<xs:schema xmlns:xs="http://www.w3.org/2001/XMLSchema"
  xmlns:tns="http://www.sbb.ch/RCMDX/Events/Defect"
  targetNamespace="http://www.sbb.ch/RCMDX/Events/Defect"
  elementFormDefault="qualified">

  <xs:include schemaLocation="../RcmDxDataTypes.xsd" />

  <xs:element name="Defect">
    <xs:complexType>
      <xs:sequence>
        <xs:element name="PossibleDefectNames" type="xs:string" minOccurs="0" maxOccurs="unbounded" />
        <xs:element name="PossibleClassifications" type="xs:string" minOccurs="0" maxOccurs="unbounded" />
        <xs:element name="PossibleValidationResults" type="xs:string" minOccurs="0" maxOccurs="unbounded" />
      </xs:sequence>
      <xs:attribute name="Classification" type="xs:string" use="required" />
      <xs:attribute name="DefectName" type="xs:string" use="required" />
      <xs:attribute name="Details" type="xs:string" use="required" />
      <xs:attribute name="Parameter1Name" type="xs:string" />
      <xs:attribute name="Parameter1Value" type="xs:string" />
      <xs:attribute name="Parameter2Name" type="xs:string" />
      <xs:attribute name="Parameter2Value" type="xs:string" />
      <xs:attribute name="Parameter3Name" type="xs:string" />
      <xs:attribute name="Parameter3Value" type="xs:string" />
      <xs:attribute name="ID" type="tns:UUID" use="required" />
    </xs:complexType>
  </xs:element>
</xs:schema>

Events Generic

<?xml version="1.0" encoding="UTF-8"?>
<xs:schema xmlns:xs="http://www.w3.org/2001/XMLSchema"
  xmlns:tns="http://www.sbb.ch/RCMDX/Events/Generic"
  targetNamespace="http://www.sbb.ch/RCMDX/Events/Generic"
  elementFormDefault="qualified">

  <xs:include schemaLocation="../RcmDxDataTypes.xsd" />

  <xs:element name="Corrupt">
    <xs:complexType>
      <xs:simpleContent>
        <xs:extension base="xs:string">
          <xs:attribute name="Username" type="xs:string" use="required" />
          <xs:attribute name="ID" type="tns:UUID" use="required" />
        </xs:extension>
      </xs:simpleContent>
    </xs:complexType>
  </xs:element>

  <xs:simpleType name="ObjectConsistencyXml">
    <xs:restriction base="xs:string">
      <xs:enumeration value="Ok" />
      <xs:enumeration value="OnlyInReal" />
      <xs:enumeration value="OnlyInData" />
      <xs:enumeration value="Measured" />
    </xs:restriction>
  </xs:simpleType>

  <xs:element name="DetectedObject">
    <xs:complexType>
      <xs:sequence>
        <xs:element name="object" minOccurs="1" maxOccurs="1">
          <xs:complexType>
            <xs:simpleContent>
              <xs:extension base="xs:string">
                <xs:attribute name="Type" type="xs:string" use="required" />
                <xs:attribute name="Description" type="xs:string" use="required" />
                <xs:attribute name="ObjectConsistency" type="tns:ObjectConsistencyXml" use="required" />
              </xs:extension>
            </xs:simpleContent>
          </xs:complexType>
        </xs:element>
        <xs:element name="Reference" minOccurs="0" maxOccurs="unbounded">
          <xs:complexType>
            <xs:simpleContent>
              <xs:extension base="xs:string">
                <xs:attribute name="ReferenceSystem" type="xs:string" use="required" />
              </xs:extension>
            </xs:simpleContent>
          </xs:complexType>
        </xs:element>
        <xs:element name="ObjectAttribute" minOccurs="0" maxOccurs="unbounded">
          <xs:complexType>
            <xs:simpleContent>
              <xs:extension base="xs:string">
                <xs:attribute name="Key" type="xs:string" use="required" />
              </xs:extension>
            </xs:simpleContent>
          </xs:complexType>
        </xs:element>
      </xs:sequence>
      <xs:attribute name="ID" type="tns:UUID" use="required" />
    </xs:complexType>
  </xs:element>

  <xs:element name="LimitViolation">
    <xs:complexType>
      <xs:attribute name="TimestampMaxViolation" type="xs:long" use="required" />
      <xs:attribute name="ViolatedLimit" type="xs:string" use="required" />
      <xs:attribute name="ID" type="tns:UUID" use="required" />
    </xs:complexType>
  </xs:element>

</xs:schema>

RCM-DX Data types

<?xml version="1.0" encoding="UTF-8"?>
<xs:schema xmlns:xs="http://www.w3.org/2001/XMLSchema" elementFormDefault="qualified">
  <xs:simpleType name="restrictedString">
    <xs:restriction base="xs:string">
      <xs:minLength value="1" />
      <xs:maxLength value="512" />
      <xs:pattern value="[a-zA-Z0-9_\-\.]+" />
    </xs:restriction>
  </xs:simpleType>
  <xs:simpleType name="restrictedStringWithColon">
    <xs:restriction base="xs:string">
      <xs:minLength value="1" />
      <xs:maxLength value="512" />
      <xs:pattern value="[a-zA-Z0-9_\-\.:]+" />
    </xs:restriction>
  </xs:simpleType>
  <xs:simpleType name="restrictedID">
    <xs:restriction base="xs:ID">
      <xs:minLength value="1" />
      <xs:maxLength value="512" />
      <xs:pattern value="[a-zA-Z0-9_\-\.]+" />
    </xs:restriction>
  </xs:simpleType>
  <xs:simpleType name="restrictedIDREF">
    <xs:restriction base="xs:IDREF">
      <xs:minLength value="1" />
      <xs:maxLength value="512" />
      <xs:pattern value="[a-zA-Z0-9_\-\.]+" />
    </xs:restriction>
  </xs:simpleType>
  <xs:simpleType name="versionString">
    <xs:restriction base="xs:string">
      <xs:minLength value="1" />
      <xs:maxLength value="32" />
      <xs:pattern value="[a-zA-Z0-9_\-\.]+" />
    </xs:restriction>
  </xs:simpleType>
  <xs:simpleType name="portNumber">
    <xs:restriction base="xs:int">
      <xs:minExclusive value="0" />
      <xs:maxInclusive value="65535" />
    </xs:restriction>
  </xs:simpleType>
  <xs:simpleType name="ipAddress">
    <xs:restriction base="xs:string">
      <xs:pattern value="(([0-9]{1,3}\.){3}[0-9]{1,3})" />
    </xs:restriction>
  </xs:simpleType>
  <xs:simpleType name="network">
    <xs:restriction base="xs:string">
      <xs:pattern value="(([0-9]{1,3}\.){3}[0-9]{1,3})/[0-9]{1,2}" />
    </xs:restriction>
  </xs:simpleType>
  <xs:simpleType name="hostName">
    <xs:restriction base="xs:string">
      <xs:pattern value="(([a-zA-Z0-9]|[a-zA-Z0-9][a-zA-Z0-9\-]*[a-zA-Z0-9])\.)*([A-Za-z0-9]|[A-Za-z0-9][A-Za-z0-9\-]*[A-Za-z0-9])" />
    </xs:restriction>
  </xs:simpleType>
  <xs:simpleType name="ipAddressOrHostName">
    <xs:union memberTypes="ipAddress hostName" />
  </xs:simpleType>
  <xs:simpleType name="nonNegativeInt">
    <xs:restriction base="xs:int">
      <xs:minInclusive value="0" />
    </xs:restriction>
  </xs:simpleType>
  <xs:simpleType name="positiveInt">
    <xs:restriction base="xs:int">
      <xs:minInclusive value="1" />
    </xs:restriction>
  </xs:simpleType>
  <xs:simpleType name="positiveFloat">
    <xs:restriction base="xs:float">
      <xs:minExclusive value="0" />
    </xs:restriction>
  </xs:simpleType>
  <xs:simpleType name="positiveIntOrMinus1">
    <xs:restriction base="xs:int">
      <xs:minInclusive value="-1" />
    </xs:restriction>
  </xs:simpleType>
  <xs:simpleType name="positiveLong">
    <xs:restriction base="xs:long">
      <xs:minExclusive value="0" />
    </xs:restriction>
  </xs:simpleType>
  <xs:simpleType name="nonNegativeLong">
    <xs:restriction base="xs:long">
      <xs:minInclusive value="0" />
    </xs:restriction>
  </xs:simpleType>
  <xs:simpleType name="compressionLevel">
    <xs:restriction base="xs:integer">
      <xs:minInclusive value="0" />
      <xs:maxInclusive value="9" />
    </xs:restriction>
  </xs:simpleType>
  <xs:simpleType name="mimeType">
    <xs:restriction base="xs:string">
      <xs:pattern value="[!#$%'*+\-0-9A-Z\^_'a-z{|}~]+/[!#$%'*+\-0-9A-Z\^_'a-z{|}~]+(; *[^;]+)*" />
    </xs:restriction>
  </xs:simpleType>
  <xs:simpleType name="nonEmptyString">
    <xs:restriction base="xs:string">
      <xs:minLength value="1" />
    </xs:restriction>
  </xs:simpleType>
  <xs:simpleType name="UUID">
    <xs:restriction base="xs:string">
      <xs:pattern
        value="(urn:uuid:)?[0-9a-fA-F]{8}-[0-9a-fA-F]{4}-[0-9a-fA-F]{4}-[0-9a-fA-F]{4}-[0-9a-fA-F]{12}|\{[0-9a-fA-F]{8}-[0-9a-fA-F]{4}-[0-9a-fA-F]{4}-[0-9a-fA-F]{4}-[0-9a-fA-F]{12}\}" />
    </xs:restriction>
  </xs:simpleType>
  <xs:simpleType name="vehicleNumber">
    <xs:restriction base="xs:string">
      <xs:pattern value="[0-9]{2} [0-9]{2} [0-9]{4} [0-9]{3}-[0-9]" />
    </xs:restriction>
  </xs:simpleType>
  <xs:simpleType name="httpUrl">
    <xs:restriction base="xs:anyURI">
      <xs:pattern value="https?://.+" />
    </xs:restriction>
  </xs:simpleType>
</xs:schema>