T-Ridership: A web tool for reprogramming public transportation fleets to minimize COVID-19 transmission

Saba Imani; Majid Vahed; Shreya Satodia; Mohammad Vahed

doi:10.1016/j.softx.2023.101350

T-Ridership: A web tool for reprogramming public transportation fleets to minimize COVID-19 transmission

SoftwareX. 2023 May:22:101350. doi: 10.1016/j.softx.2023.101350. Epub 2023 Mar 9.

Authors

Saba Imani¹, Majid Vahed², Shreya Satodia¹, Mohammad Vahed³

Affiliations

¹ School of Architecture, University of Southern California, Los Angeles, CA, USA.
² Pharmaceutical Sciences Research Center, Shahid Beheshti University of Medical Sciences, Tehran, Iran.
³ Department of Clinical Pharmacy, School of Pharmacy, University of Southern California, Los Angeles, CA, USA.

Abstract

As the outbreak of novel coronavirus disease (COVID-19) continues to spread throughout the world, steps are being taken to limit the impact on public health. In the realm of infectious diseases like COVID-19, social distancing is one of the effective measures to avoid exposure to the virus and reduce its spread. Traveling on public transport can meaningfully facilitate the propagation of the transmission of infectious diseases. Accordingly, responsive actions taken by public transit agencies against risk factors can effectively limit the risk and make transit systems safe. Among the multitude of risk factors that can affect infection spread on public transport, the likelihood of exposure is a major factor that depends on the number of people riding the public transport and can be reduced by socially distanced settings. Considering that many individuals may not act in the socially optimal manner, the necessity of public transit agencies to implement measures and restrictions is vital. In this study, we present a novel web-based application, T-Ridership, based on a hybrid optimized dynamic programming inspired by neural networks algorithm to optimize public transit for safety with respect to COVID-19. Two main steps are taken in the analysis through Metropolitan Transportation Authority (MTA): detecting high-density stations by input data normalization, and then, using these results, the T-Ridership tool automatically determines optimal station order to avoid overcrowded transit vehicles. Effectively our proposed web tool helps public transit to be safe to ride under risk of infections by reducing the density of riders on public transit vehicles as well as trip duration. These results can be used in expanding on and improving policy in public transit, to better plan the scheduled time of trains and buses in a way that prevents high-volume human contact, increases social distance, and reduces the possibility of disease transmission (available at:http://t-ridership.com and GitHub at: https://github.com/Imani-Saba/TRidership).

Keywords: COVID-19; Machine learning; Neural networks; Public health; Public transportation planning; Social distancing; Travel restrictions; Web-based application.