Research Impact on Practice

Deriving All Passenger Flows in a Railway NEtwork

from Ticket Sales Data

FUll Paper

IAROR 2011

Rome, Italy

16-18 Feb 2011

Presentation held

at both conferences

 

EXTENDED ABSTRACT

ORBEL 2011

GHENT, BELGIUM

10-11 Feb 2011

MT-ITS 2011, Leuven, BELGIUM, 22-24 JUN 2011

Calculation of Realistic Station Capacity by PlatformING Feasibility Checks

Expected Passenger Travel Time

for Schedule Evaluation & optimization

Presentation held

at conference

 

AUTOMATED, PASSENGER TIME OPTIMAL, ROBUST TIMETABLING

USING MIXED INTEGER PROGRAMMING

IWHIR 2011, SHENZHEN & HONG KONG, CHINA, 19-22 JUL 2011

Presentation held

at conference

 

ABSTRACT

The research papers below are the result of R&D carried out by Peter Sels for Logically Yours for its client Infrabel. The papers were all written by Peter Sels. They were peer reviewed by Pieter Vansteenwegen, Thijs Dewilde and Dirk Cattrysse. They were subsequently submitted to or/and published in conferences proceedings or in journals. For conferences the presentation slides are reproduced here as well.


In short, the impact on practice of this research is that Infrabel now possesses 2 working prototypes of tools: RhinoCeros and Leopard.


















  1. (1)What does RhinoCeros compute? It

  2. -reads the route and timing (action minima and action current supplements) of all passenger trains departing in a given hour in the current timetable

  3. -keeps the routing and station halting pattern but changes the timing supplements, resulting in an optimized timetable

  4. -minimizes the criterium of expected passenger time in practice

  5. -produces an optimized  timetable that, for Belgium, saves about 3.8% of average journey time on average over all passengers!


  1. (2)Why is RhinoCeros unique? It

  2. -(unlike *all* other MILP cyclic timetabling research) gives a guarantee that a  feasible timetable is found (given that in the input timetable, capacity constraints are not exceeded),

  3. -gives a result that is optimal in expected passenger time in practice, which includes assumed delays, based on measured delays from the current timetable,

  4. -considers *all* potential transfers (>17000 for Belgium, and not only a list of transfers that are assumed to be important),

  5. -can retime a timetable and subsequently reroute passengers and do this iteratively, further reducing  expected passenger time,

  6. -contains an integer linear programming model that scales up to a full country like Belgium, which operates  around 200 passenger trains per hour,

  7. -corrects all errors against minimum times on riding, dwelling actions and on headway constraints,

  8. -(unlike other research) only needs two hours of computation time (for Belgium) to obtain this result.

















  1. (3) What does Leopard compute? It

  2. -produces , for a user selected station, an  IN-route, a platform and an OUT-route assignment for as many trains as possible. This is the goal function required by Infrabel.

  3. -also considers and avoids all route conflicts in the IN grid as well as OUT grid, so no train couples will come into conflict when the platform-plan is put into operation,

  4. -gives a graphical dynamic SVG output showing each train occupation time interval on their platform,

  5. -corrects train pair conflicts that are present in the current station platforming plans.


  1. (4)Why is Leopard unique? It

  2. -shows the status for every train pair using crossing or equal routes within an interval of 5 minutes: overlapping, non-robustly close, or robustly separated (of course the optimized solution never contains overlaps),

  3. -(unlike some other research) calculates the platforming and routing plan for medium stations in less than a second.
















  1. (5)What principles are core to Logically Yours?

  2. (a) for Attitude:

In common programming one must consider all cases that can occur in reality.

In optimal programming one must  return the best solution that is mathematically possible.

Compromises in both of the above are common.

We take no compromises in either domain.

The reason is the knowledge and  experience that such compromises *always* backfire when the client is confronted with reality.

So, quite simply put, we will do what it takes to get the optimal solution in all cases. No cutting corners! Yes, it is more effort, but the client’s product goes from unreliable to reliable and from just ok to giving solutions that cannot be beaten.























  1. (b)for Bridge:

Apart from the above ‘attitude’, we also believe in the usefulness of bridging two worlds:

  1. - the academic world (university cooperation as a PhD, academic publishing world, interaction at conferences) and

  2. -real practice by working for a client with real challenges: the Belgian infrastructure management company Infrabel.


Knowledge about the insights, approaches and results of academic peers leads to a quicker understanding of what approach should or shouldn’t be taken. It also makes it easier to see holes in the current state of the art. Discussion with peers also helps to find conceptual errors or new approaches.


Knowledge about the industrial practice, allows one to get the necessary input data quicker, avoids academic overly simple or overly complex models of reality and makes that the prototype tool is directly integrated in the client’s tool flow and way of working.


“Researchers in mathematical optimization should grasp the currently available momentum and opportunities in the railway industry by not focusing too much on theoretical results, but by going for real world applications of their models and techniques. The latter will lead to a win-win situation, both for the researchers and for the railway industry.”


Professor Alberto Caprara et al. (2006)




(c) for Conclusive, Closure, Courage:

Re-search, by definition, can be improved upon forever. In industry however, some needs are urgent or have a limited window of opportunity. This opportunity should be taken - at the right time - or too much valuable research will remain unapplied.



Because of the use of our above 3 core principles, (a,b,c) , we believe that we can achieve the balance the client wishes to go for between:

  their goal of ‘maximising impact on practice‘

  their goal of ‘minimising research to useful integrated product time’.

Presentation held

at conference & aT Phd Seminar 26 MArch, 2013

FULL PAPER

ABSTRACT

Research Publications

IAROR 2013 Conference,Copenhagen, Denmark, 13-15 May 2013

Extended Abstract

 

SHORT PAPER

 

A Passenger Knock-On Delay Model for Timetable Optimisation

PAPER

FULL PAPER

PUBLISHED ONLINE ON 17/1/2014

BY ELSEVIER  JOURNAL

TRANSPORTATION RESEARCH, PART B


Volume 61, March 2014, Pages 55–72

http://www.sciencedirect.com/science/article/pii/S0191261514000058

http://authors.elsevier.com/offprints/TRB1196/41cb9677601828708356947222e5732f

THE TRAIN PLATFORMING PROBLEM:

THE INFRASTRUCTURE MANAGEMENT COMPANY PERSPECTIVE

Enjoy!

Automatically and Quickly Planning

Platform and Route of Trains in Railway Stations

with a case study of MEchelen Station

International Journal Conference on Civil and Transport Engineering

(ISAET'14, 1-2 JANUARY 2014)

Presentation Held at CONFERENCE,

3 December, 2013

Automatically and Quickly Planning Platform

and Route of Trains in Railway Stations

Presentation Held at IFORS, BARCELONA CONFERENCE,

14 JULY, 2014

optimal TEmporal Spreading of Alternative Trains

in order to Minimise Passenger Travel Time in Practice

Presented At THE IAROR CONFERENCE, Tokyo, 23-26 MArch, 2015

ABSTRACT

Towards a Better Train Timetable for Denmark

Reducing Total Expected Passenger Time

Presented at the CASPT Conference, Rotterdam, 19-23 July, 2015

Practical Macroscopic EvaluatIon and Comparison

of railway  Timetables

Presented at THE EWGT CONFERENCE, Delft, 14-16 July, 2015

Full Paper

ExtendED ABSTRACT

MT_ITS 2013, DRESDEN CONFERENCE, 2-4 December 2013

Full Paper

Short Course Presented At THE IAROR CONFERENCE, Tokyo, 23 MArch, 2015

Timetabling

Platforming

Reducing the Passenger Travel Time in Practice
by the Automated Construction of a Robust Railway Timetable

ABSTRACT

Published by The Journal : Transportation Research Part B

DOI: 10.1016/j.trb.2015.12.007

http://www.sciencedirect.com/science/article/pii/S0191261515002684

Full Paper

Full Paper

Presentation Held at CONFERENCE,

23 March, 2015

FULL PAPER

ExtendED ABSTRACT

Presentation

Also presented at Euro2015, Glasgow, UK, July 12-15, 2015

by Pieter

Presentation

Automated Platforming & Routing of Trains

in All Belgian Railway Stations

Accepted by The Journal : Expert Systems With Applications

Full Paper

Light rail, tram and bus timetabling

Minimising Passenger Time in Practice

Accepted by The Conference : Tristan. Oranjestad, Aruba,  13-17 June, 2016

Extended Abstract

Full PhD Text, Including the most important papers above

Presentation PDF

Leuven, 03/05/2016

PhD Thesis

May 2016

FLyer PDF 3/05/2016

Presentation VIDEO

Leuven, 03/05/2016