Examining regularity in vehicular traffic through Bluetooth scanner data: Is the daily commuter the regular road user?

Regularity is an important property of individual travel behaviour. The ability to measure it enables advances in behaviour modelling, mobility prediction and commuter analytics. The majority of work in this space focuses on regularity in public transport ridership using different sources of big data. Yet, another important consideration is the regularity of vehicular traffic and road users in urban contexts. In this paper we analyse road user travel using Bluetooth scanner data obtained for three contextually different urban regions over a period of six months. Sequences of weekly travel events - discrete repeatable behaviour units for each road user were identified by using the unique Bluetooth Mac-addresses of enabled devices (e.g. vehicle navigation system, mobile phone). We then present a cluster analysis of traveller typologies, sensitive to both the frequency and order the travel events. Overall, we find that the regular daily commute accounts for a small percentage of vehicle travel. Our results reveal typologies of road user travel behaviour that differ across the city and reflect localised land use characteristics. This is important for traffic regulation and parking planning.