Page 61 - Proceeding of Atrans Young Researcher's Forum 2019_Neat

“Transportation for A Better Life:
Smart Mobility for Now and Then”

23 August 2019, Bangkok, Thailand

Experiment on mobility survey using smartphone in Hanoi, Vietnam
Topic number: 04 Paper Identification number: AYRF2019-008
1
2
Minh Hieu NGUYEN , Jimmy ARMOOGUM , Cédric GARCIA 3

1 Université Paris-Est, IFSTTAR/AME/DEST (France)
Faculty of Transport and Economics, University of Transport and Communications (Vietnam)
E-mail: minh-hieu.nguyen@ifsttar.fr or hieunm@utc.edu.vn

2 Université Paris-Est, IFSTTAR/AME/DEST (France)
E-mail : jimmy.armoogum@ifsttar.fr

3 Université Paris-Est, IFSTTAR/AME/DEST (France)
E-mail : cedric.garcia@ifsttar.fr

Abstract
Conventional travel survey techniques like face-to-face interview, computer-assisted telephone
interview, (e-)mail have big drawbacks including heavy burden on participants, low-response rate and high
cost, not to mention missing short and unimportant trips. Surveys using GPS are trendy because positioning
data are objective, numerous and accurate in terms of space and time. This paper presents the preliminary
results of an experiment on observing mobility using smartphone in Hanoi, Vietnam. Overall, 63 participants
provided travel data along with the corresponding ground truth via the app named TRavelVU developed by
Trivector (Sweden). They showed their happiness and interest in participating in the survey. Besides, some
technical issues and privacy concern were reported. The authors developed a deterministic method on the
basis of speed profiles to distinguish between walk, bike and motorized segments. The overall accuracy of
87% was encouraging. The inclusion of motorcycle was the main reason for the misclassification between
bike and motorized modes. This paper has contributed to the geographical diversity of both smartphone-based
mobility survey and mode detection fields that has greatly focused on developed countries with basic modes
of walk, bike, car, bus/train, metro.

Keywords: GPS, mobility survey, mode detection, smartphone, rule-based

1. Introduction
Mobility data are essential for travel GPS-based mobility surveys have been frequently
forecast. In the past, self-reported household surveys carried out in developed countries such as France,
using pen and paper or telephone played the major Australia, US, New Zealand [5]–[8] but not in the
role [1]. Unfortunately, these techniques have big Global South. In this paper, initial results of an
shortcomings including heavy burden on experiment on mobility survey using smartphone in
participants, low-response rate and high cost, not to Hanoi (Vietnam) were presented. The following of
mention missing short and unimportant trips [1], [2]. this paper was structured into four main sections.
The use of GPS can limits the mentioned-above The first reviewed the literature of travel
issues because its data are objective with spatio- investigations employing GPS technology and
temporal details. Initially, on-board GPS devices methods to detect transportation mode. The next
were utilized to take advantage of electricity power. described the experiment whilst the third section
Later, lightweight and wearable loggers have been showed how to use rules to detect travel modes from
used [3]. Now, smartphones equipped GPS sensors GPS data. Conclusion part closed this paper.
are the main tool due to its ability to visualize points
on map thus allowing and supporting users to 2. Literature review
validate data directly on the same application. 2.1 Smartphone-based mobility survey
Besides, persons hardly forget their phone; Smartphones have been deployed in the
therefore, continuous collection is achieved [4]. cases of Singapore [4], China [9], Canada [10] and
New Zealand [8].

36