Vehicle Traffic and Flood Monitoring with Reroute System Using Bayesian Networks Analysis Melchizedek I. Alipio * , Jess Ross R. Bayanay † , Alex O. Casantusan ‡ and Abigail A. Dequeros § Electronics Engineering Department, Malayan Colleges Laguna, Pulo Diezmo, Cabuyao City 4025 Philippines Email: * m.i.alipio@ieee.org, † jrrbayanay@live.mcl.edu.ph, ‡ aocasantusan@live.mcl.edu.ph and § aadequeros@live.mcl.edu.ph Abstract—Heavy vehicle traffic and flooded areas are problems experienced on roads because of unimproved road infrastructures and environmental deviations. These factors affect vehicle drivers negatively as they contribute to stress, health problems, and wastefulness of time. This study developed a system called ArRoad that monitors and analyzes vehicle traffic and flooded areas using network of sensors and real-time image processing which then predicts and visualizes possible alternative rerouting paths using machine learning. Water level sensor nodes are used to monitor the flooded areas while real-time video images from cameras are processed to extract the vehicle volume on the streets. A Bayesian Network is generated from the water level sensors and image processing data which provides possible reroute areas to avoid traffic congestion and flooded areas. All data are sent to a cloud platform through the Internet that can be accessed through a mobile user interface. This mobile user application provides information about the condition of the streets and possible reroute maps to users. The accuracy of the system is tested by actual implementation on a specific road. Results showed minimum accessing delay from using the ArRoad to navigate in rerouted paths to prevent impassable roads due to heavy traffic and flood. If effect, it lessens the amount of time experienced by drivers from heavy traffic condition and flooded streets which then improves the quality of life by preventing waste of resources such as time and money. Index Terms—Bayesian Network; Image processing; Internet of Things: Machine learning; Sensors; Vehicle traffic I. I NTRODUCTION Internet of Things (IoT) has become the center of idea in developing infrastructures for it is manageable to reinforce applications on interacting smart devices. IoT has allowed these smart devices to be used in developing operations in a network that is significantly convenient for daily activities where its services are widened through the Internet [1]. Many issues have been solved through the development of these smart devices technology especially on road traffic congestion and flood monitoring matters. This use cases have been utilized globally where smart devices have been set up in such a way that it is built with decision support systems through Wireless Sensor Networks (WSN) [2], [3]. Based on different studies on traffic congestion and flood monitoring, the relief of these problems needs to improve on open source implementations and real-time optimization. The Philippines is one of the countries suffering from both heavy floods and traffic congestions. This problem would eventually worsen if not given effective solutions that can affect commuters badly and worst, the countrys economy as a whole. The main goal of this work is to develop a system for local drivers that reliably monitors and analyzes vehicle traffic and flood conditions and performs predictive analysis of possible alternative rerouting paths to avoid congested and flooded road areas. This paper makes the following contributions: (1) develop a sensor network that monitors flood level; (2) generate a predictive rerouting path using Bayesian Network based from real-time image processing and flood level data; and (3) develop a user mobile interface that served as the viewing access in monitoring vehicle traffic, flooded areas and possible reroutes. The rest of the paper is organized as follows. Section II presents the related work. In Section III, we discuss an overview of the ArRoad system and the corresponding testbed deployment. We present our results in Section IV. Finally, Section V concludes the paper. II. RELATED WORKS Previous works presented various solutions in monitoring vehicle road traffic and flood level integrated with IoT to provide alternative paths for the drivers and public commuters. In [4], an Intelligent Transport System (ITS) is developed to serve as a solution to the waste of energy caused by traffic con- gestion and inefficient traffic. Two major concepts are needed to be considered to make the ITS possible; the availability of real-time accurate traffic flow data and the interface of the traffic flow forecast model. To address the solution to the problem with the real-time collection of data, WSNs were applied. It would serve as self-organizing networks that have a large number of nodes integrating information collection, data processing, and wireless communication. Through WSNs, it was said that the system can collect information about the real-time behavior of traffic that could support the analyzation of the mathematical algorithm. However, it was not clear how data were processed and accessed with actual testbed set-up. On the other hand, [5] have demonstrated a project that showed the effectivity and accuracy of a wireless sensor device as a traffic and flood monitoring scheme. Ultrasonic Rangefinder is used to perform traffic sensing measurements with the highest possible accuracy. Infrared sensors were used to allow traffic monitoring on a higher number of traffic lanes, and the infrared thermopiles are for simultaneous measures the temperature in the field of view. The suggested sensor contains an ultrasonic rangefinder and multiple passive in- frared temperature sensors. Because of the tremendously low absolute distance measurement error required by the system, 2017 IEEE 6th Global Conference on Consumer Electronics (GCCE 2017) 978-1-5090-4045-2/17/$31.00 ©2017 IEEE