Congestion continues to increase as the city reopens in the wake of the COVID-19 pandemic. If public transit is to be part of the solution to increasing congestion, then system performance matters. This is the topic of interest to University of Southern California researchers Cyrus Shahabi, Yao Yi Chiang, and Luan Tran, and the subject of their recently released report, “Large-scale and Long-term Forecasting of Performance Measurement of Public Transportation Systems.”  

In this Caltrans-funded project, the researchers sought to develop long-term and large-scale forecasting tools, including the development of a bus arrival time web dashboard for transit riders, for the entire Los Angeles Metropolitan Area (LAMA) public transportation system. Their approach utilized both spatial statistical methods and machine learning, using data from the University of Southern California Archived Transportation Data Management System (ADMS) to predict bus arrivals at the city-level. The forecasting tool also provides benefits to the transit agency. By quantifying the efficiency (e.g., delays) of bus lines, transportation agencies can schedule buses more strategically and supply riders with better information on arrival and departure times. They can also do so with greater accuracy and efficiency. The research also contributes to better trip planning, policymaking, and vehicle management. One of the greatest potential benefits is the ability of multiple agencies to undertake long-term forecasting on a truly massive scale – enough to lead to comprehensive and lasting changes for Angelenos.   

Shahabi, Chiang, and Tran hope to continue refining their approach to include data from several transportation modes and event information for major events where an influx of traffic might be expected, including football games. This work will be critical as the City of L.A.’s traffic steadily creeps back to pre-pandemic levels and disruptions from in-person events return. Perhaps in the not-so-distant future, studies such as this one will lead to a public transit system as flexible and dynamic as the citizens who rely on it.