Conference Papers 2017

Seminar
Day 1 (4 Oct 2017), Session 1, Big Data, 11:30 - 13:30

Status
Accepted, documents submitted

Submitted by / Abstract owner
Chris Wroe

Authors
Chris Wroe, Telefonica, Stephen Rutherford, Jacobs, Paul Hanson, AECOM

Short abstract
This paper introduces and evaluates methods used to refresh TfL’s Strategic Transport model using mobile network data fused with secondary datasets.

Abstract
The Strategic Analysis department at Transport for London (TfL) is responsible for maintaining and updating the transport models which are used to forecast future demand in planning transport strategy for London. Through Project EDMOND, TfL are working with AECOM, Jacobs and Telefonica to update these models using data from the O2 mobile network fused with a variety of other datasets relating to transport in London.
The mobile network data is processed to identify trips where users have moved between two distinct parts of the cell network. Longer term behavioural trends are used to infer trip purpose, and the clustering and geospatial patterns of the data will be used to begin to infer mode. Particular areas of innovation from previous work will include the use of a probabilistic method to distinguish mode and route, calibrated using various secondary data sources including road traffic counts, vehicle GPS traces, station entry counts and smartcard ticketing data.
The paper will introduce the work done to process the mobile network data and to fuse it with other sources to create a complete picture of transport in London. Different data sources will be reviewed and evaluated in terms of their compatibility with mobile network data. The paper will also suggest the future opportunities for how mobile network data might be used to enhance transport modelling in cities.
This abstract is submitted subject to approval by Transport for London. It is anticipated that this will be obtained by March 2017. More detail may also be added to the abstract at this stage.