Can Smartphone Location Data at the Point Level be Used to Estimate Traffic Volumes?: A Methodological Evaluation

Verduzco-Torres, J. R. and Raturi, V. (2023) Can Smartphone Location Data at the Point Level be Used to Estimate Traffic Volumes?: A Methodological Evaluation. 12th International Conference on Geographic Information Science (GIScience 2023), Leeds, UK, 14-19 Sept 2023.

Abstract

The effective development of transport planning strategies relies on timely and comprehensive urban travel information. Manual or automatic traffic counts form a key input into many models but are available only for a limited set of locations and at a low frequency. The emergence of digital footprint data, particularly location data at the point level obtained from smartphones, has raised the possibility of replacing or supplementing traditional analytical methods and sources. However, the suitability of the data for these purposes is currently poorly understood. This study aims to assess the accuracy of estimating traffic counts and travel direction volumes by utilizing a sample of high-resolution smartphone location data. Counts will be derived and compared to 44 manual count points published by the Department of Transport in 2019 at various locations in and around Glasgow. The location data is provided by a private organization, Huq, which comprises 22 million records from 19,000 unique users for the same year in the study area. Two main challenges are addressed in the paper. The first is effectively extracting travel activity from the location data while mitigating the noise stemming from non-travel activities and imprecise or sparse location estimates. To address this, we construct and test a series of systematic buffers around traffic count points based on the characteristics of the road network and the immediate built environment. The second challenge concerns the heterogeneity in the quality and volume of the data produced by app users. We further examine whether it is possible to refine the estimates by inferring individual trips and the direction of travel based on the patterns of the location data. Preliminary results indicate that the proposed methodology explains between 36% and 70% of the variability of the manual traffic counts in a regression framework. This study establishes a methodological precedent and suggests directions for future research. Potential avenues include investigating additional years or cities, incorporating more temporally detailed traffic count data, integrating CCTV-based counts, evaluating alternative location data sources, and focusing on active travel modes.

• Item Type: Conference or Workshop Item
• Status: Published
• Refereed: No
• Glasgow Author(s) Enlighten ID: Verduzco-Torres, Dr Jose Rafael and Raturi, Dr Varun
• Authors: Verduzco-Torres, J. R., and Raturi, V.
• College/School: College of Social Sciences > School of Social and Political Sciences > Urban Studies
• Copyright Holders: Copyright © 2023 The Authors
• Publisher Policy: Reproduced with the permission of the authors

Related URLs

• Organisation