Privacy-preserving Methods for Traffic Data Collection and Analysis

Research Lead: Wenlong Jin

UC Campus(es): UC Irvine

Problem Statement: Traditional methods for data collection, such as the National Household Travel Survey, focus on trips by a small sample of either travelers, locations, or times. With the prevalence of GPS devices and smartphones, big transportation data from more travelers and locations over longer timespans are more readily available and can substantially help to improve the management, planning, and design of transportation systems. However, travelers, private companies, and public agencies are reluctant to share such data due to privacy concerns.

Project Description: This project will develop a new privacy-preserving method for collecting and analyzing traffic data. This method is based on a new framework for transportation system analysis, in which a network is considered a single entity, and trips are tracked in a relative space with respect to the remaining distance to individual travelers’ destinations. Such data are sufficient for characterizing traffic dynamics but without revealing Personally Identifiable Location Information. This method works for either a city road network or freeway corridors, as well as for multimodal trips. The project will systematically calibrate and validate the new method and will discuss the policy implications for data collection and analysis for California’s traffic systems.

Status: In Progress

Budget: $87,000

Project Partner(s): Caltrans