Conference Papers 2018

Seminar
Day 2 (11 Oct 2018), Session 5, Modelling Time, Speed and Price Using Analytics, 11:15 - 12:45

Status
Accepted, documents submitted

Submitted by / Abstract owner
Siamak Khorgami

Authors
(1) Anwar Yusuf, AECOM
(2) Siamak Khorgami, AECOM*
(3) Callum Hale, Leicestershire County Council (LCC)
(4) Bin Deng (Debbie), AECOM

Short abstract
This paper reports on the outcome of a collaborative research funded by AECOM and Leicestershire County Council setting out a practical approach to assess the appropriateness of use of Google Maps journey time data in transport planning applications.

Abstract
Access to a robust journey time database is crucial for monitoring congestion levels and understanding reliability of journey times on the road network. Traditionally, such data are usually obtained through Moving Car Observer (MCO) surveys, camera observations from Automatic Number-Plate Recognition systems (ANPR) or from using/processing Global Positioning System (GPS) derived from journey times of vehicles.

Google Maps has provided crowd-sourced journey time information to its consumers for many years, and this source is of potential use for transport modelling and planning (as well as other uses such as transport policy development, journey time reliability/network resilience monitoring), through products such as the Google Maps Distance Matrix API (1) and Google Maps Directions API (2).

Compared to other sources of journey time data, Google Maps journey time data have wider geographical coverage, higher sample size, it captures the day-to-day variability of travel time, and has potential time and cost savings for data collection and processing.

There is only limited evidence on the verification of Google Maps journey time data in transport planning and modelling literature. This paper discusses an approach to assess the appropriateness of use of Google Maps journey data in transport planning and modelling through defining a case study in England. A representative sample of routes are used to cover different road types, representing a mixture of radial, arterial and local road links with varying traffic condition and differing level of congestions.

In particular, this study discusses: (a) how and to what extent Google Maps journey time data are comparable in quality to the existing/conventional journey time data sources; (b) the appropriateness of Google Maps journey time data for the development of strategic transport models and microsimulation models; and (c) the strength and weaknesses of Google Maps and conventional data sets in relation to data collection and processing.
Finally, this paper will explore other potential applications of Google Maps journey time data, such as journey time reliability and network resilience analysis, developing congestion indicators and monitoring congestion level, evaluation and monitoring schemes, validation of transport model network speeds, real time data and predictive modelling, and traffic management.

(1) The Google Maps Distance Matrix API provides with distance and durations for multiple destinations and transport modes. https://developers.google.com/maps/documentation/distance-matrix/

(2) The Google Maps Directions API “return multi-part directions for a series of waypoints”, and “directions for several modes of transportation are available”. https://developers.google.com/maps/documentation/directions/

The authors acknowledge the contributions and efforts of Alex Gray and Sonny Tolofari of Leicestershire County Council in supporting and funding the data collection, supporting the data analysis, and providing insightful oversight for this research project.

Programme committee
Data

Topic
New Mobility patterns, Digitisation and Mobility as a Service