"IoT-Based Vehicle Monitoring and Driver Assistance System Framework for Safety and Smart Fleet Management"

Abstract

Curbing road accidents havealways been one of the utmost priority of nations worldwide. In Malaysia, the Traffic Investigation and Enforcement Department reported that Malaysia’s total number of road accidents haveincreased from 373,071 to 533,875 in the last decade. One of the significant causes of road accidentsis the driver’s behaviors. However, to regulate drivers’ behavior by the enforcement team or fleet operatorsischallenging, especially for heavy vehicles. In our research, we have proposed the Internet of Things (IoT) scalability framework and its’ emerging technologies to monitor and alert driver’s behavioral and driving patterns to reduceroad accidents. To prove this work, we have implementeda lane tracking,and iris detection algorithm, to monitor and alert the driver’s behavior when the vehicle sways away from the lane, and to detect if the driver is feeling drowsy. We implemented electronic devices such as cameras, a global positioning system module, a global system communication module, and a microcontroller as the hardware for an intelligent system in the vehicle. We also appliedface recognition for person identification using the same in-vehicle camera and recorded the working duration for authentication and operation health monitoring. With the GPS module, we monitored and alerted against permissible vehicle’s speed accordingly. We integrated IoT on the system for the fleet centre to monitor and alert the driver’s behavioral activities in real-time through the user access portal. We have validated it successfully on Malaysian roads. The outcome of this pilot project ensuresthe safety of drivers, public road users, and passengers. The impact of this framework leads to a new regulation by the government agencies towards merit and demerit system, real-time fleet monitoring of intelligent transportation systems, and socio-economy such as cheaper health premiums. The big data can be used to predict the driver’s behavioral in the future.