USE OF BLUETOOTH TRACKING TECHNOLOGY IN TRANSPORT MODELLING



USE OF BLUETOOTH TRACKING TECHNOLOGY IN TRANSPORT MODELLING

Authors

Tony Dichev, Transport for London, Paul Jackson, Sky High Technology Ltd

Description

The technique of using the unique Bluetooth signatures being emitted from vehicles to validate the TfL ONE traffic model is described. 32 million signatures were recorded during the survey period and outputs were generated to validate the model.

Abstract

The TfL ONE traffic model was extended in 2013 and a need to validate the new model was identified. The relatively new technique of using the unique Bluetooth signatures emitting from vehicles was adopted as an alternative to traditional methods.

The primary objective of the study was to generate observed origin-destination vehicle trip matrices for various time periods using cordon and screenline sites which could be optionally Furnessed to control totals and to use these to validate the modelled movements produced by the extended model.

Temporary roadside sensors were installed for 10 days in March 2013 to capture the unique Bluetooth signatures emitting from vehicles at over 110 locations around London. Automatic Traffic Count data were collected concurrently for a two week period.

32 million raw Bluetooth signatures were recorded during the survey period which after de-duplication resulted in 28m valid recordings. 740,000 unique signatures were identified. An interface was created to allow trip matrices to be generated and factored and to analyse trip routing between sites. 13m valid journeys were identified and 1.8m unique routes were observed. Outputs were generated which allowed the predictions of the model to be compared with the observed movements.

To our knowledge, this is the largest temporary Bluetooth data collection exercise undertaken and is a unique case of using such data to validate a traffic model of this scale.

The study illustrates that data of this type can be collected and analysed in large volumes across a city network. The data is at the time of submitting this Abstract still being used to validate the model, but conclusions from this will be available for the Conference.

Publisher

Association for European Transport