A Machine Intelligent Cloud Based Framework to Monitor Driving Behavior of Vehicles Using Internet of Things

Abstract

Monitoring driving behavior of vehicles is  a very important area of research in the field of Intelligent  Transportation Systems (ITSs). As vehicles are rising  tremendously on the roads of cities, it is very much  essential to track the driving behavior of a vehicle to reduce  the accidents in the cities. The drivers who are not driving  properly, for example if a driver is continuously  performing sudden acceleration, sudden deceleration,  sudden left turn, sudden right turn, etc. then these types of  vehicles need to be monitored and detected. In this work, a  machine intelligent cloud-based framework is proposed to  monitor the driving behavior of vehicles in the city using  internet of things (IoTs). The vehicles are installed with an  on-board unit which is responsible for collecting all  readings from the sensors and sending these readings to  cloud using IoT for detection of driving patterns. Here, the  cloud is deployed with a supervised machine intelligent  model that is responsible for identifying the driving pattern  after receiving the readings from a vehicle. The model is  selected by conducting training and testing over a standard  dataset. The performance of the framework is tested using  Python tool. From the results, it is found that Random  Forest (RF) model performs better in identifying the  accurate vehicle behavior with highest classification  accuracy (CA).