Crowdsourcing Big Data from Smartphone Apps for Transportation Research: Role of Statistics and Challenges

• Activity Number: 241
• Type: Topic Contributed
• Date/Time: Monday, August 10, 2015 : 2:00 PM to 3:50 PM
• Sponsor: Government Statistics Section
• Author(s): Feng Guo* and Arash Mirzaei* and Elaine Murakami* and Bianica Pires* and Tianjia Tang*
• Companies: Virginia Tech and North Central Texas Council of Governments and Federal Highway Administration and Virginia Tech and Federal Highway Administration
• Keywords: Crowd-sourcing ; Big Data ; Smartphone Apps ; Transportation Statistics

Abstract

The proliferation of smartphone provides a great opportunity to collect travel and road infrastructure information from both road users and the general population. Compared to traditional survey-based method, the cellphone approach does not require specialized instruments and can provide continuously data feeding from large sample size. The cellphone approach also makes the collection possible for these traditionally difficult-to-collect data items such as exposure from alternative transportation modes. Federal Highway Administration, state department of transportation, and local agencies have initiated various projects and programs to explore the related procedures, processes, and data. While these explorations reveal great opportunities and potentials, significant challenges are also being identified. Privacy issue, sampling control, variability and noise, data weighting and expansion et al are still formidable roadblocks to the acceptance of the entire community. This panel consisted of experts from federal, state and local government agencies and academics will provide an overview of the current state-of-practice and how traditional statistical practices can be empowered t