What Big-Data Does Not Tell Us. What Can We Learn from Travel Surveys for Bus and Light Rail in the Netherlands

What Big-Data Does Not Tell Us. What Can We Learn from Travel Surveys for Bus and Light Rail in the Netherlands


Marcel Bus, Panteia BV, Wouter Kuhlman, TU Delft, Jan Kiel, Panteia BV


The paper presents new insights in travel behaviour of users of bus, tram and metro in the Netherlands. The results are relevant for an different (international) stakeholders.


In 2008 the Dutch public transport (PT) operators introduced a new digital ticket system (chip card). This new system allows to collect a lot of data about travelling flows by bus and light-rail, such as distance and time travelled between two bus stops or at what time-of-day people travel. Despite this rich source, it is difficult to retrieve travel behaviour on aspects such as the purpose of a journey or the real origin and destination. Another drawback of the collected data is, that it concerns sensitive information. For privacy reasons the data is deleted after one and a half year. Due to this, we turned to a source that was used before the introduction of the chip card, to see whether this source could help to improve our knowledge on travel behaviour by bus, metro and light rail.

Until 2008, public transport by bus in the Netherlands used a national ticket system. This system has been used for more than 20 years. With one ticket you could travel with different public transport operators. The revenues were returned to the PT operators. The distribution of these revenues was based on a large survey. This survey contains more information than the chip-card. One needs to think of purpose, real origin-destination or acces and egress. The complete survey contains about two million respondents for the period 1984-2009. To get an idea of the richness of this source, we performed an analysis for the period 2003-2009, which consists of more than half a million respondents.

The aim of the analysis was to see whether the chip-card data could be enriched by using the travel surveys. This may serve different purposes, such as providing rules of thumb or improving transport models. This will help different stakeholders such as PT operators, regional PT concession authorities, governments and consultants to underpin different policy measures, as well as to increase their knowledge of public transport by bus and light-rail. The combination of analog collected data and big data may help to get deeper insights.

The surveys were analysed by Panteia for different aspects, such as the average trip distance to a bus-stop, depending on the level of urbanisation. Also, we were able to analyse travel patterns per hour per purpose. The information has been transformed into distribution functions. What is important is, that over the years these functions appeared to be rather stable. It allows us to use the information from the surveys to enrich the data retrieved from the chip-card. In our paper we will present some of the results. These results are relevant for both Dutch and international stakeholders, as both data and functions of this kind are hard to find, if available at all. The results help to improve the (international) knowledge on the use of public transport by bus, metro and light-rail.


Association for European Transport