COVID-19’s Effect on Transportation: Developing a Public COVID-19 Data Dashboard

Overview

The COVID-19 outbreak has dramatically changed travel behavior in cities across the world. With changed travel demand, economic activity, and social-distancing/stay-at-home policies, transportation systems have experienced an unprecedented shift in demand and usage. Since the start of the pandemic, the C2SMART research team has been collecting data and investigating the impact of COVID-19 on mobility and sociability, including:

Passenger travel and freight traffic trends
Mode shift and usage based on various policies
Effect of social distancing policies on transit use and emissions
Sidewalk, crosswalk, and intersection crowd density
Effect of COVID-19 Policies on Transportation Systems

This project was made possible by collaboration from University of Washington (Jingxing Wang, Yanyan Chen, Sai Sarath Chandra Pavuluri Venkata, Xuegang (Jeff) Ban); Rutgers University (Chaekuk Na, Hani Nassif); Old Dominion University (Hong Yang, Qingyu Ma); Cornell University (H. Oliver Gao, Mohammad Tayarani); and Ulteig (Abdullah Kurkcu).

Research Objectives

Leveraging open data from multiple sources, this project developed a Sociability Data Dashboard which features both traditional and innovative techniques, such as data mining and visualization, agent-based traffic simulation model and real-time computer vision technique, to help researchers and transportation authorities understand and observe the impact of the pandemic on transportation.

To investigate crowd density and the effectiveness of social distancing strategies, C2SMART researchers have introduced a low-cost, AI-driven big data acquisition framework leveraging hundreds of traffic cameras along with a deep learning-based video processing method.

Object detection and distance approximation between pedestrian pairs are applied to traffic camera videos at multiple NYC and Seattle locations to analyze local social distancing patterns. This sociability board shows some examples of the application.