Travel Mode Detection Based on GPS Raw Data Collected by Smartphones: A Systematic Review of the Existing Methodologies

Abstract

Abstract:Over the past couple of decades, global positioning system (GPS) technology has been utilized to collect large-scale data from travel surveys. As the precise spatiotemporal characteristics of travel could be provided by GPS devices, the issues of traditional travel survey, such as misreporting and non-response, could be addressed. Considering the defects of dedicated GPS devices (e.g., the need for a large sum of money to buy devices, forgetfulness in the taking of devices to collect data, limiting of the sample size because of the number of devices, etc.), and the fact that the smartphone is becoming one of the necessities of life, there is an opportunity for smartphones to replace dedicated GPS devices. Although several general reviews of GPS travel survey and GPS data-processing methods have been written, a systematic review of smartphone-based GPS data collection and travel mode detection has not be made.. The studies were collected from six databases. The purpose of this review is to critically evaluate the current literature on the existing methodologies of travel mode detection based on GPS raw data collected by smartphones. Meanwhile, according to a systematic comparison of the different methods of data preprocessing for travel mode detection, this paper details the strengths and weaknesses of the existing methods. Furthermore, it is a very important step towards developing methodologies and applications for GPS raw data collected by smartphones.

Obesity and the Built Environment: A Reappraisal

“…Acknowledging that an individual’s home neighborhood cannot capture the full extent of health‐related BE exposure, researchers turned to more advanced GPS technologies to track movements in space and time . Cumulative mobility over a given time period is the basis for constructing AS areas .…”

“…The literature examining whether BE features captured in an individual’s AS are more predictive of health than home‐centric‐defined features is relatively nascent . In some cases, AS‐defined food and structural environment features have been shown to be more predictive of health behaviors than those captured in the home neighborhood.…”

“…Accordingly, such BE measurement tools have become increasingly more sophisticated using wearable GPS devices, often augmented with travel diaries, travel logs, Google street views, or other mobile technologies . These approaches allow the capture of near instantaneous measures of exposure to the BE not only around the home location but also at work, school, and along travel routes . Mobile technologies have become increasingly atomized, allowing researchers to capture the minutiae of daily activities in time and space, sometimes at time resolutions of 10 seconds or less.…”

Developing a passive GPS tracking system to study long-term travel behavior

“…Today, the focus has shifted away from GPS loggers towards smartphone applications (Cottrill et al (2013)), due to their easier administration and the development of automatic methods for detecting transport mode based on GPS data. Most of these methods are based on machine learning techniques and they often integrate GPS data with data from other smartphone sensors such as accelerometers (two related reviews are Wu et al (2016) and Nikolic and Bierlaire (2017)). Most of today's smartphone-based tracking systems use a prompted recall approach.…”

A Random Forest Model for Travel Mode Identification Based on Mobile Phone Signaling Data

“…The main principle of the probability-based algorithm is to define a possible matching area as the center of the matching point according to the positioning accuracy of the track points, take the road section within the area as the possible matching road section and calculate its matching probability to determine the best matching road. In addition to the above map-matching algorithms, matching algorithms using complex mathematical theories have emerged in recent years, such as the Bayesian inference matching algorithm, Kalman filtering matching algorithm, fuzzy logic matching algorithm, and matching algorithm based on convex optimization [29][30][31][32].…”