Upload
others
View
4
Download
0
Embed Size (px)
Citation preview
2nd UIC RailTopoModel and railML®
Conference
Towards a Universal Topology Model for Railways and Data Exchange Format for Infrastructure
Session A:Results from UIC’s ERIM activity: UIC RailTopoModel
UIC Paris; April 8th, 2014
Session A – Page 2 2nd UIC RailTopoModel and railML® Conference; Paris, April 8th, 2014
Agenda
1 Welcome and keynote speeches
2 Results from UIC’s ERIM activity: UIC RailTopoModel
3 The railML® 3 Prototype for Railway topology
4 The usage of railML® for the RINF project of ERA
5 Project plan and next steps
Session A – Page 3 2nd UIC RailTopoModel and railML® Conference; Paris, April 8th, 2014
Reminder from Sept 2013: Our challenge
Based on the positive results of UIC feasibility study
> Propose to the railway community§ A powerful and strong object model
§ A structured and evolutionary exchange format
internalpartners
Session A – Page 4 2nd UIC RailTopoModel and railML® Conference; Paris, April 8th, 2014
Reminder from Sept 2013: our commitment
> Finalize the foundation of UIC RailTopoModel
> Build the first layers of railML 3
> Propose the first bricks for railMLtool box
> Build a railML translator for RINF
Session A – Page 5 2nd UIC RailTopoModel and railML® Conference; Paris, April 8th, 2014
UIC RailTopoModel Basics: Users
The main aim of the UIC RailTopoModel is- to provide a structure for the description of a rail network- that supports the needs of the two main categories of rail network users
- Asset Management (mainly use the positioning functions)
- Traffic Management (mainly use the routing capabilities)
ObjectObject ReferenceAxis
ReferenceAxis
Session A – Page 6 2nd UIC RailTopoModel and railML® Conference; Paris, April 8th, 2014
UIC RailTopoModel Basics: Referencing I
X, YX, Y
λ, φλ, φ
Km 31,515Km 31,515Km 30 + 214mKm 30 + 214m
0,77540,7754
30
Many ways to reference an event, in space or along a line :
Session A – Page 7 2nd UIC RailTopoModel and railML® Conference; Paris, April 8th, 2014
We chose the percentage of the “edge” (trail, section of line, …) as the “Master” system for storing and transmitting position data, because
- of its independence from Spatial Referencing system (Works in schematics as well as geographic),- of its independence from the units (miles, km, m, degrees…), physical references or measuring
tapes – they are usable at any level- Every other system can easily fall back to this one- Every other system can be deduced from it
Coordinate transformations functions will be offered to translate from this system to more user-friendly referencing (GPS, XYZ, Mileage, Mileposts…)
UIC RailTopoModel Basics: Referencing II
0,77540,7754
30
0
1
Session A – Page 8 2nd UIC RailTopoModel and railML® Conference; Paris, April 8th, 2014
A1
C0
D0
We want to model a network.
Every element of the network is related to others.These relations happens at one of their extremities (0 or 1) These links may be navigable or not …
UIC RailTopoModel Basics: Relations
A
C
DF
C1
D1
Session A – Page 9 2nd UIC RailTopoModel and railML® Conference; Paris, April 8th, 2014
The users does not see the network in the same way:
Micro : Detail infrastructure
Meso : connections between stops, sidings and junctions
Macro : connections between major cities
UIC RailTopoModel Basics: Multilevel I
Session A – Page 10 2nd UIC RailTopoModel and railML® Conference; Paris, April 8th, 2014
Those visions of the network describe the same network!
We have to ensure: - Data consistency
The network should really model the same reality, whichever the level
- Ease of data maintenanceThe data must be managed at the level where it is collected – 1! time.It should then be manipulated to fit its uses – Multiple usages.
Example: Gauge reduction: happens at a specific location - managed as a located event
Usage: Gauge capability on the segment = minimum gauge for this segment
UIC RailTopoModel Basics: Multilevel II
Session A – Page 11 2nd UIC RailTopoModel and railML® Conference; Paris, April 8th, 2014
Micro :
Meso :
Macro :
The aggregation mechanics ensures that - network definition stays coherent from one level to the other and- information can be transmitted from one level to another.
UIC RailTopoModel Basics: Aggregation
Session A – Page 12 2nd UIC RailTopoModel and railML® Conference; Paris, April 8th, 2014
UIC RailTopoModel Basics: Common PathWe defined 3 “common” aggregation levels, along one “aggregation path”;
The model allows the definition of any number of other levelsAlong this path or another.
Micro
“Micro for Meso”
Meso
Macro
“Micro for Micro+”“Micro for Micro+”
“Micro+ for Meso”“Micro+ for Meso”
Micro+
Disaggregation
Aggregation
“Micro for X“Micro for X
“X for Y”“X for Y”
X
“Y for Macro”“Y for Macro”
Y
AA
B
Virtual Level
Persistent LevelZ
“Y for Z”“Y for Z”“Meso for Macro”“Meso for Macro”
Common Path
MicroMicro
Session A – Page 13 2nd UIC RailTopoModel and railML® Conference; Paris, April 8th, 2014
UIC RailTopoModel Basics: Objects IWe have identified three types of objects/events which occur on the railway network :
Punctual object : Signals, boundaries, beacons, …
Linear object : Route, slope profile, speed profile, ballast renewal, platform edge…
Areal object : Track circuit, catenary zone, Station, bridge…
Session A – Page 14 2nd UIC RailTopoModel and railML® Conference; Paris, April 8th, 2014
UIC RailTopoModel Basics: Objects II
Punctual object : Transmitted as a position on a line
Linear object : Transmitted as an ordered list of objects
Areal object : Transmitted as an unordered list of objects, forming a subnetwork
s
1
2 3
Session A – Page 15 2nd UIC RailTopoModel and railML® Conference; Paris, April 8th, 2014
Next steps after establishment of the UIC RailTopoModel
Concept:
>Topology = kernel (carrier) of railway objects and data.
Metaphors of the Christmas tree and the flowerbud.
Future common extensions to the UIC RailTopoModel
Session A – Page 16 2nd UIC RailTopoModel and railML® Conference; Paris, April 8th, 2014
Christmas tree metaphor
Kernel = treestem + branches= network topology
(macro, meso, micro)
carrier of
garlands, birds, bells and balls
= decoration
(all types of objects)
Session A – Page 17 2nd UIC RailTopoModel and railML® Conference; Paris, April 8th, 2014
Flower bud metaphor
Kernel = bud; origin of leavesJellow cone = generic
extension on the budLeaves = application
domains
• Railway topology = kernel• Future kernel extensions• ERTMS, timetabling, asset
management, etc.= application domains
Session A – Page 18 2nd UIC RailTopoModel and railML® Conference; Paris, April 8th, 2014
1. Always present (at a certain level) on railways:à a set of safety rules and their methods
2. Always valid for each object:à a set of coordinates for the location of
objects, network-nodes, area demarcations…
3. Always relevant for any data item or object:à time interval of existence
4. Any set of generic
(gauges, gradient, ballast type…)
interlocking
Locations 3D
Time validity
Utilization restrictions
Some possible generic extensions on the topology
Session A – Page 19 2nd UIC RailTopoModel and railML® Conference; Paris, April 8th, 2014
Generic extensions:Concentric circles around the kernel:
Session A – Page 20 2nd UIC RailTopoModel and railML® Conference; Paris, April 8th, 2014
Norwegian Experiences: JernbaneverketTools and data available for development
National asset DB
Session A – Page 21 2nd UIC RailTopoModel and railML® Conference; Paris, April 8th, 2014
Norwegian Experiences: JernbaneverketNeed for an integration hub
National infrastructureassets db(RDBMS)
Integration hub(RDF format:Flat structure)
SQL RDF
RDFSQL -> RDF
RDF
RDF -> RailML
xxx -> RDF
yyy -> RDFxxx
yyy
RDF
RDF
RailML
RDF
RDF
RDF -> SQL
RDF -> csv
RailML
RINF
RailML
Session A – Page 22 2nd UIC RailTopoModel and railML® Conference; Paris, April 8th, 2014
Swiss experiences: BLS / SBBStatus Quo
Beginning in 2005, SBB has created a unified topology database called UNOserving several operational business cases such as timetable-planning or train-disposition. It's base concept is a Node/Edge Graph Model where other data such as signalisation can be easily linked. Topology information changes over time and therefore the concept and implementation requires a solid time dimension approach, which has been implemented.
The following analysis has been made having the concepts of UNO in mind which has been industry proven over the last 7 years.
Gleisknoten GKn
Gleispunkt GP
Gleiskante GKa
Gleisanfang/ende GAENetwork-Reference
Signaling System in Inventory
Network-Topology
Source: Markus Giger/http://de.wikipedia.org
Session A – Page 23 2nd UIC RailTopoModel and railML® Conference; Paris, April 8th, 2014
Swiss experiences: BLS / SBBPossible Domains
• Exchange format SBB / BLS / regulation authority / third parties
• Asset-Management with a common topology• ONE topology serving several business cases
Session A – Page 24 2nd UIC RailTopoModel and railML® Conference; Paris, April 8th, 2014
Swiss experiences: BLS / SBB
Compatibility ChecküUIC RailTopoModel could be (to be verified) generated from existingSwiss applications. UNO alreadyuses a node/edge graph-model.
üIntrinsic positioning system as base reference system to avoidanomalies in linear positioning.
üBased upon one MICRO level, several MESO - and if required evenmore MACRO levels - can beimplemented.
Open Issues•Time dimension / integrity•Status information / integrity•Distinct definitions•Univeral identifiers•Define what specific objectsshould be modeled on whichlevel (by example)
Image source: http://www.masternewmedia.org,
http://wikipedia.org
Session A – Page 25 2nd UIC RailTopoModel and railML® Conference; Paris, April 8th, 2014
Swiss experiences: BLS / SBBSummary
• In Switzerland there is a need for a common exchange format
• Even with some important open issues, the UIC RailTopoModel has a high level of maturity
• Relevant stakeholders should participate on defining a standard and must finally commit to it
Image source: http://www.processcollection.de
Session A – Page 26 2nd UIC RailTopoModel and railML® Conference; Paris, April 8th, 2014
Targets• RailTopoModel and railML as a
building block of IT-Strategy?• Contribution to development of
UIC RailTopoModel
Context• Urgent need: improve quality of combined
permanent way & track geometry data • Introduce topology into construction and
maintenance centered business processes• Promote integrated view of railwayActions
• Analyse and choose relevant datasets• Understand UIC RailTopoModel• Derive and implement a relational table design• Import existing data• Implement example functionality Results
• Working pilot on the line level• Model related feedback• Integration of track level data in preparation
Austrian experiences: ÖBB InfraUIC RailTopoModel and ÖBB data models I
Session A – Page 27 2nd UIC RailTopoModel and railML® Conference; Paris, April 8th, 2014
Context• Urgent need: improve quality of combined
permanent way & track geometry data
Testbed for compatibility check• Permanent way dataset contains topology
information on track level• Navigation in user interface requires topology
information on line level
New technical regulation of track layout• Schwerpunkttrassierung – barycenter track layout• Safety considerations require precise switch data in
track layout environmentà data quality improvement necessary
Austrian experiences: ÖBB InfraUIC RailTopoModel and ÖBB data models II
Session A – Page 28 2nd UIC RailTopoModel and railML® Conference; Paris, April 8th, 2014
Short digression• Topology can be used to improve
documentation quality systematically
Austrian experiences: ÖBB InfraUIC RailTopoModel and ÖBB data models III
Session A – Page 29 2nd UIC RailTopoModel and railML® Conference; Paris, April 8th, 2014
End of short digression
Austrian experiences: ÖBB InfraUIC RailTopoModel and ÖBB data models IV
Session A – Page 30 2nd UIC RailTopoModel and railML® Conference; Paris, April 8th, 2014
Analyse and choose relevant datasets
ÖBB dataset Content UIC RailTopoModel
IOS Line and track designation framework
MACRO and MESO
iGleislagemessung Configuration data for high performance measurement car
MICRO / MESO
iOberbau Permanent way MICRO (NANO)
Austrian experiences: ÖBB InfraActions I
Session A – Page 31 2nd UIC RailTopoModel and railML® Conference; Paris, April 8th, 2014
Understanding UIC RailTopoModel
Micro
Meso
Macro
Standard PathÖBB “full-size”
Nano
Macro 2
Macro
Meso
Micro
ÖBB “RailTopoModel”
Test
Macro
MicroPermanent way detailed view (one switch has three connection points)Permanent way; Timetable
Routeplanning for measurement car
Standard lines
Aggregated lines(e.g. two lines on one railway embankment)
Austrian experiences: ÖBB InfraActions II
Session A – Page 32 2nd UIC RailTopoModel and railML® Conference; Paris, April 8th, 2014
Implement example functionality
Functionality Description Concept of UIC RailTopoModel
Station, stop and junction navigation
Present adjacent stations, stops and junctions and allow direct jump to one of the neighbouring sections
NetElement and Relation
jurisdictional areas Show appropriate service region and service center
Network
Alternative line names Allow user to choose line naming system and present correct line code and name
NetEvent
Austrian experiences: ÖBB InfraActions III
Session A – Page 33 2nd UIC RailTopoModel and railML® Conference; Paris, April 8th, 2014
Working piloton line level
• Jurisdictionalareas
• Flexible namingsystems
• Navigation functionalityfor userinterface
Austrian experiences: ÖBB InfraResults
Session A – Page 34 2nd UIC RailTopoModel and railML® Conference; Paris, April 8th, 2014
Model related feedback
Issue Description ActionReferences Allow references in
RailTopoModel to backlink to existing datasets
Extension of model
Vertical navigation How to navigate between different levels of a layered network (Micro, Meso, Macro)
Clarify preferred modelling approach
Using existing standards
e.g. OGC GML (Geography Markup Language)
Discussion
Lifecycle How to model the lifecycle of a layered network
Discussion
Austrian experiences: ÖBB InfraResults II
Session A – Page 35 2nd UIC RailTopoModel and railML® Conference; Paris, April 8th, 2014
• UIC RailTopoModel has the potential to become the relevant referencemodel for topology components of integrated railway IT – Systems
• UIC RailTopoModel will allow to streamline interfaces betweeninfrastructure and operating business processes
• UIC RailTopoModel will allow to rectify existing systems and will be agood starting point for new developments
• railML is perfectly positioned to implement UIC RailTopoModel ascentral part of a comprehensive exchange format for all aspects ofrailways
Austrian experiences: ÖBB InfraConclusions
Session A – Page 36 2nd UIC RailTopoModel and railML® Conference; Paris, April 8th, 2014
Business Uses case: European Corridors
> Challenge:§ 6 corridors, crossing more than 10 countries§ Each Corridor has build its own
Information System/GIS. § All concerned IM’s have to provide fine
and up-to-date infrastructure descriptionof each segment of network.
§ Some IM’s might have to feed 3 or 4 different Corridors IS
> Proposal:§ railML 3 shall be the standard format
for all topology and infrastructure data exchanges between IMs and Corridor IS