Content area
Full Text
Abstract- Mexico City has the most elevated traffic congestion in the world according to a recent report carried out by TomTom navigation products. A driver in Mexico City could spend up to 59% extra time trapped in traffic during a normal day. This paper applies a cluster analysis to find out the heavy traffic zones using massive data from the Social Navigation Network Waze. This zonification is fundamental for urban planning and transportation management, and may help develop accurate prediction models for heavy traffic.
Keywords: Geo-spatial Data Analysis, Clustering, Vehicular Traffic, Data Mining
1. Introduction
The rapid urban growth of Mexico City has made it one of the most crowded in the world with a population close to twenty million people. As a result, the Metropolitan area is in constant expansion, forcing thousands of people to use their vehicles daily. Thus, this year the levels of air pollutants have exceeded the maximum exposure limits established by the World Health Organization (WHO) [1]. The heavy traffic in Mexico City causes a daily loss of 3.3 million man hours in traffic, which costs 33 billion pesos according to a study of The Mexican Institute for Competitiveness (IMC)[2].
An important characteristic of developing cities is the heavy traffic zones, where main roads converge and generate a large amount of traffic. The detection of such traffic zones allows seriously analyzing causes and proposals for realistic solutions. In the past, this problem has largely been approached by specialists with knowledge of the city. Nowadays, information technologies enable the detection of heavy traffic zones using massive and real time data. We refer to sensor, video and GPS devices [3].
The increasing use of smart devices with a GPS has promoted the development of different social navigation networks such as: Waze [4], Google Maps [5] and Inrix Traffic [6], etc. In Mexico City, the most popular of these networks is Waze, which collects data from user devices and allows them to report events and traffic levels, generating a real-time snapshot of the traffic situation.
This paper analyzes heavy traffic zones in Mexico City using data from Waze. In the light of Data Mining algorithms we propose a zonification of the metropolitan area based on massive quantities of user-reports. Section 2,...