Research: Mezuro Data shows outstanding correlation to survey OViN

Mezuro is measuring a vast amount of 1,2 billion travelling days a year. What is the origination, what is the destination, at what time does one leave, travelling by car or train, etc. This supplies us with a huge amount of (aggregated) mobility information which can be used for several mobility issues. For example the measurement of the effects of a variety of actions (mobility management, Beter Benutten (Better Use), road construction). This could also serve as input/replacement of current models based on surveys. Naturally it is interesting to compare these big data results with manually executed surveys by OViN.

Comparison with Surveys OViN by students Business Informatics UU

Jens van Langen and Johan Meppeling, interns at Mezuro/Decisio and both graduating at Universiteit Utrecht in Business Informatics, compared Mezuro data/information with OViN. OViN is short for Onderzoek Verplaatsingen in Nederland (research displacements in the Netherlands) and this is executed by the CBS (central bureau for statistics). In this study people are being asked via survey to record for a day where they are going that day. Time of departure, as is manually filled in in OViN surveys, was compared to times of departure originating from Mezuro data. This showed major correlation (Pearson correlation coefficient is 0,977). This is graphically shown below.


enquetes OViN correlatie

More accurate and specific travelling behaviour data

We can conclude that the Mezuro data correlates outstandingly well with the surveys held by the CBS for journeys of over 10 kilometres. In fact, Mezuro data is even more accurate because mobile data gives a larger sample size ( a vast 1.2 billion travelling days a year) than surveys can ever give. With this we can draw conclusions about mobility behaviour in specific regions (e.g. nearby Utrecht) specific relations (e.g. Amsterdam – Almere) or specific hours (e.g. comparison of summer and winter, days in the week, morning or evening rush hour). With this information we can make traffic- and infrastructure policy more effective and efficient. This makes taking the right measures in the right time and place and monitoring the effects of traffic scheduling and policy a lot simpler.

For more information: Wim Steenbakkers