Spatiotemporal Analysis of Traffic Accidents and Traffic Congestion Along Commonwealth Avenue Using Crowdsourced Waze Traffic Data

Abstract. With the growing demand for transportation and traffic management in Metro Manila, innovation in these industries is essential. Crowdsourced traffic data, particularly from smartphones, has emerged as an accurate and reliable source. This study utilized crowdsourced traffic data from Waze, through the Waze for Cities (WFC) program, to explore the relationship between traffic accidents and congestion. Focusing on Commonwealth Avenue, an accident-prone highway, and by applying Kernel Density Estimation (KDE), six areas of interest (AOIs) were identified. These areas align with more general accident-prone zones identified by the Metro Manila Development Authority (MMDA) through the Metro Manila Accident Recording and Analysis System (MMARAS) Annual Report. The study examined the behavior of traffic jams in these areas, considering variables such as congestion extent, additional time delay, and travel speed. Results showed that the presence of accidents significantly increased the speed difference between regular and jammed conditions, as indicated by the Mann-Whitney U test at a 5% significance level, except in two areas of interest (AOI) near Quezon Memorial Circle, which confirms the potential of using Waze data in assessing traffic conditions. From the findings, it is suggested that Waze data be integrated with traditional data sources used by government agencies for more effective traffic management, enhancing the accuracy and responsiveness of traffic assessments, interventions, and policy implementations.