Punctuality prognosis with OnTime
Ensuring a high quality of service starts in early planning stages and needs powerful software tools. To asses the quality of timetables Schweizerische Bundesbahnen (SBB) ordered the implementation of the so called "Stabilitätsmodul". It completes SBB´s software toolbox by adding a software suite for timetable stability analysis.
In order to gauge the quality of traffic achievable in a timetable or, more generally, in an operating concept, it is necessary to predict how potential delays will propagate through the railway network. Random variables are used to model the types of delays arising with the aid of distribution or density functions.
Using analytical procedures to compute the propagation of delays on major railway networks yields sizeable advantages over Monte Carlo simulations concerning the size of the area under investigation, the speed of analysis and the quality of results.
This approach expands the possibilities of timetable quality assessment compared to other simulation software:
- it is possible to analyse timetables for entire regional or national networks instead of only a network sector. So for the first time it is possible to quantify the complex networkwide effects of timetable alterations.
- Subjecting delay propagation to analytical mapping allows dependable computations to be speedily achieved even for large networks. Unlike in Monte-Carlo simulation, where concrete realisations of the delay situation are generated for each simulation run, analytical methods directly process the distribution functions of the underlying random variables. Events that in practice lead to alterations in a delay are mapped by manipulating the distribution functions in analytical methods.
The analytical approach makes it possible to calculate the punctuality of networkwide timetables in minutes rather than hours/days. - The short time for the timetable simulation allows for new possibilities in the process of timetable design and timetable evaluation.
- The possibility to reproduce an actual disturbance level makes it possible to calibrate the model and provide accurate prognoses of punctuality.
The basic input needed for any assessment of timetable quality in OnTime are
- Timetable and accompanying operational information (connections, turn-arounds, ...)
- Infrastructure properties (headways, ...)
- Primary delays to describe the level of disturbance a railway system has to deal with
The implementation of the timetable stability tool OnTime was done by cooperation of trafIT solutions gmbh, Zurich, and VIA Consulting & Development GmbH, Aachen. It is based on proven and tested methods developed at the Institute of transport science (VIA) of RWTH Aachen University.
Apart from SBB it is used by Infrabel (Belgium), DB InfraGo (Germany), ProRail (NL), Trenitalia (IT), by the Institute for Transport Planning and Systems of the ETH Zurich, the Institute of Transport Science at RWTH Aachen University and in several consulting mandates.
Further Information
Literature:
For literature on some fundamentals on OnTime see the Journal of Rail Transport Planning & Management:
Stochastic modelling of delay propagation in large networks
Proc. of the 5th International Seminar on Railway Operations Modelling and Analysis, Copenhagen,
13.-15.05.2013 of the International Association of Railway Operations Research (IAROR):
OnTime-TimetableStabilityAnalysis.pdf
Combining Quality of Service and Operations for Timetable Evaluation: "Kombinierte betriebliche und verkehrliche Bewertung von Bahn-Angeboten" (german)
Eisenbahntechnische Rundschau (ETR), (2018) Nr. 7+8, S. 26-29
IRSA 2017: talk/paper at the 1. International Railway Symposium Aachen: Kombinierte betriebliche und verkehrliche Bewertung von Bahn-Angeboten (german)
COMPRAIL 2022: Dan Burkolter, Burkhard Franke, Bernhard Seybold: "IMPROVING RAIL NETWORK SIMULATIONS WITH DISCRETE DISTRIBUTIONS IN ONTIME"
Releases
November 2024
OnTime 5.6
- various technical improvements (reduced memory usage when displaying keyfigures and geoView) and improvements to the GUI
- new optional extension: the line map.
The line map display the directional delay/punctuality of all trains of a line in the geographic view.
July 2024
OnTime 5.5
- corrections to the extension Availability to simulate large disruptions (Monte-Carlo - style) in combination with OnTimes distribution-based delay calculation
- filtering of trains in the GTT
- GTT-style according to Cronos (SBB-tool for analysis of operational data)
- measuring delays and punctuality with regard to commercial arrival and departure times (instead of operational times) in the Keyfigure and the Map View
April 2024
OnTime 5.4
- new additional result file: summary statistics for every train
- small improvement to the graph-view: grouping of rules for better understanding
- the OnTimeRunner exports all of the new extended output files
- changes for the import of Viriato-timetables at Infrabel
March 2024
OnTime 5.3
- new module: track occupation
- new features in the TimetableManager: delete trains, copy or shift trains, add stops
- an extended graph summary is provided after every calculation
- extended output of the conflict-indicators
- improvements to the multilingual support
September 2023
OnTime 5.2
- improvements in the DistributionView: easier selection of trains, permanent display of 0
- improvements in the Link-Generator
- SBB: new chromatic graph in the GTT
January 2023
OnTime 5.1
- Platform conflicts: A new class of conflicts considered in the calculation.
- Speedup of graph-creation: Reduction of the first part of calculation by about 30%.
- In the discrete calculation mode the length-dependant primary delay is now truly proportional to the section length (and not an approximation)
- Included a Java Runtime Engine (OpenJRE from https://Adoptium.net) in the OnTime package. It is no longer necessary to have Java installed, this is a change to the system requirements.
- SBB: Simulation of parts of a network: new features and new input file for SBB
March 2022
OnTime 5.0
- Redesign of the calculation kernel, new calculation mode using discrete modelling of distributions
- Shorter calculation times and increased accuracy
- New View: Distribution viewer showing the delay distribution for every train at every moment
March 2021
OnTime 4.3
- Improvement of the calculation kernel
- increased accuracy of the calculation
- updated the import of NeTS-railML (SBB)
January - May 2020
OnTime 4.2.1 - 4.2.6
- reworking of the calculation kernel
- reworking of the additional result files and creation of new output-file: postgraph to better retrace and understand the calculation
- improvements and performance enhancements when displaying the results
May 2019
OnTime 4.2
-
New specific headway mode "Mindestzugfolgezeiten" for DB
-
Additional module: The Timetable Analyzer, visualizing the number of trains in a timetable scenario
- Additional analysis view: Section Analysis to analyse the change of delays on the sections from stop to stop
November 2018
OnTime 4.1
-
usage of DB-speed restrictions (Tages-La)
-
Improvements in GUI and analysis
April 2018
OnTime 4.0
-
migration of the programming environment to Eclipse 4.7 and update to Java 8
-
GUI - improvements:
-
files are arranged in groups and coloured according to their file type
-
redesign and redistribution of the settings
-
inclusion of more features in the settings (like speed restrictions, departure delay)
-
redesign of the railML-Manager. Settings for timetable management can be defined and saved.
-
integrated captions in graphics
-
improvements to the localisation of conflicts and delay propagation
-
french and dutch/flemish verson of OnTime
Autumn 2017
OnTime 3.1
-
New functionality: Module Availability to simulate the effects of infrastructur availability and of disruptions on punctuality.
-
consistent internationalization: OnTime is available in english, german and italian and is ready for any other language
Jan 2017
OnTime 3.0
Modifications to Input files
- import of Viriato timetable railML 2.2
- autocompletion of missing station/junction headways
- new output: the cause-effect file for an easier analysis of delay causes and delay propagation
- display of number of tracks in the GeoView
- the trainPathUsage now affects connections and turn-arounds as well as conflicts
- Availability of new module: Train Diagram
May 2016
OnTime 2.9
Modifications to Input files
-
speed restrictions (#OnTime-SPEED_RESTRICTIONS): the track declaration becomes optional
-
Extension of the file format for conflicting routes in junctions
-
reworked the testing of input files in the graph-view
-
improved analysis of the scenario in the Indicators
Modifications to Output files
-
Improvement of the activities file. The .act file now records every incoming and outgoing activity.
-
ProRail-specific: changes to the DONS-converter. The generated headways now fit the blocking time model
User-defined color-scale for results
New Connection View to analyze the quality of connections
Various UI-improvements
July 2015
OnTime 2.8
- extended functions for the preprocessing of timetables:
- automatic generation of turn-arounds and train coupling and sharing
- generation of a useful trainNumber for uncompleted timetables (long-term / strategic timetables)
May 2015
OnTime 2.7.2
- Improved Import of DONS-data for ProRail. Automatic generation of railML-timetable, connections, turnarounds and headways.
- Extension of the file format for Primary Delays Stopping
- Extension of the file format for conflicting routes in stations
Dec 2014
OnTime 2.7
- probability attribute of constuction works
- Import of DONS-timetables for ProRail
June 2014
OnTime 2.6.6
- Import of infrabel timetable files based on railML 2.0
Oct 2013
OnTime 2.6
- Kapazitätsschätzer: capacity estimation according to UIC 406.
Powerful UIC Leaflet 406 compliant compression
approach, accounting for lines and nodes, assessment of available and consumed capacity - Consideration of fuzzy trains. When analysing future schedules, the schedules of some trains - especially freight trains - are not known in detail. Only the demand for a number of train pathes can be predicted. OnTime can combine the scheduled trains of a timetable and the predicted demand.
Mar 2013
OnTime 2.5
- improvement of the calculation: multi-core calculation is standard,
visualization of calculation progress - extended output: logging of effective primary and secondary delays
Dec 2012
OnTime 2.4
- Visualization of results as graphic timetable. Delay distribution of each train is shown in a GTT, making it possible to analysis the delay evolution and propagation per train, section and time.
Jul 2012
OnTime 2.3
Mar 2012
OnTime 2.2
Jan 2012
OnTime 2.1
Dec 2011
OnTime 2.0