Inferring Traffic Patterns of Dhaka City: A Spatio-Temporal Analysis Over a Year

Abstract

Numerous large cities’ sustainability and livability are severely hampered by traffic congestion. To create efficient measures to relieve this congestion, it is vital to comprehend the spatiotemporal patterns of this congestion. Our study addresses the scarcity of long-term spatiotemporal analysis for traffic congestion in developing megacities, focusing on Dhaka, the fifth most congested city. Utilizing big data analytics with a sample of 350,400 records from Google Maps over a year, we identify temporal patterns, spatial distribution, and recurrent congestion patterns. Through extensive image processing, peak hours, congestion variations, inter-zone relationships, and causes of extreme congestion are analyzed. Dhaka is divided into ten zones, revealing distinct congestion patterns with variations between weekends and weekdays. These findings offer crucial insights for urban planning, traffic control, and infrastructure development in rapidly expanding megacities, contributing to the alleviation of congestion and enhancement of sustainability and livability.