Dynamic Public Transport Data

[Vahid Moghani]

Hi all,

I am a researcher at Erasmus School of Economics. I am starting a project studying the effect of access to (night) public transport on risky behaviour such as alcohol abuse and unprotected sex.

For this purpose, the missing piece of puzzle is a dynamic dataset for large cities in the Netherlands (including Amsterdadm, Rotterdam, Den Haag, Utrecht, Eindhoven, Groningen, Tilburg, Almere, Breda, Nijmegen, Leiden, and Maastricht). The data I need must include all the bus, metro, and tram lanes, their start and end dates, their schedule and intermediate stations, and their temporary disruptions/cancellations. The data ideally need to go back in time till 2006. However, a dataset containing parts of 2006-2017 would still be useful.

I have been searching for the data, however, the ones I came across are either real-time or goes back to a few months/years ago for most of the cities (such as ndovloket). I was in touch with some transport companies which they claim they have the data but it requires a lot of work and they need financing for that.

I am wondering if there is any source or tricks that I can construct the dataset myself, or somewhere I can request it. I am sorry if this is a topic you already discussed, I have not yet learned Dutch.

 

Warm regards,

 

Vahid Moghani
Ph.D. Candidate
Erasmus School of Economics (ESE)

Tinbergen Institute (TI)

Erasmus Centre for Health Economics Rotterdam (EsCHER)

W   http://vahid-moghani.com/

Visiting Address T-Building / Room T18-04 / Burgemeester Oudlaan 50 / 3062 PA Rotterdam

Postal Address Postbus 17383000 / DR Rotterdam

[Stefan de Konink]

Hi Vahid,

From your message it does not become clear why you are searching
specifically for the 2006-2017 period. I can be very clear there is no
realtime information available before 2009. The standardisation of the
first realtime interface was done on in 2009-03-12. openOV started in
December 2009, at that time the maximum availablity was realtime data for
Dutch Railways. The reporting on incidents that has been used in the past
exists since 2008. Model Informatieprofiel Openbaar Vervoer (MIPOV) 2008.

So my best bet if you want to recover historic cancellations it would be
using the freedom of information act (WOB) and request the MIPOV reports.

For the period 2013 and onwards, historic realtime data is available and
can be combined into a prediction if a service was available. Either by
validating there was actually a cancellation message, or establishing that
realtime data would have been provided in other cases. Integrating this
set will require cpu time, a lot. If you can get away with just knowing
the bus did not drive, while it was scheduled it might be easier.

For the data that is available, the process is simple at NDOVloket,
provide the academic credentials. For the size of this dataset there will
be shipping costs...

Stefan

[Vahid Moghani]

Hi Stefan,

Thank you very much for the information! I believe this information solves the obstacles to a large extent.

The reason for going back in time comes from the fact that other datasets start much earlier than 2013. I have access to health surveys and medication data of individuals from 2006. Ideally, I want to look at the long-term effects in addition to the short-term consequences as well.

I can get in touch with NDOVlocket to start with the later years, and if things started to work out, I look further into the years before 2013.

For the period after 2013, I agree that having real-time data might be enough to establish both the actual and planned timetables. Otherwise, I can try to merge the real-time data with datasets on cancellation messages or planned schedules.

This Google Group is a great initiative! Thanks for your help!

Warm regards,

Vahid