DATAFITS: A HETEROGENEOUS DATA FUSION FRAMEWORK FOR TRAFFIC AND INCIDENT PREDICTION

Abstract

In order to create a complete dataset, this study presents DataFITS (Data Fusion on Intelligent Transportation System), an open-source system that gathers and fuses traffic-related data from several sources. Our hypothesis is that traffic models may benefit from improved information coverage and quality thanks to a heterogeneous data fusion architecture, which would boost the effectiveness and dependability of ITS systems. Two applications that made use of event categorization and traffic estimate models confirmed our hypothesis. For nine months, DataFITS gathered four different kinds of data from seven different sources and combined them into a spatiotemporal domain. While incident categorization utilized the k-nearest neighbors (k-NN) method with Dynamic Time Warping (DTW) and Wasserstein metric as distance measurements, traffic estimation models used polynomial regression and descriptive statistics. The findings show that by fusing data, DataFITS was able to enhance information quality for up to 40% of all roads and dramatically expand road coverage by 137%. While incident classification reached 90% accuracy on binary tasks (incident or non-incident) and about 80% on categorizing three distinct categories of events (accident, congestion, and non-incident), traffic estimate earned an R2 score of 0.91 using a polynomial regression model.