OBD-II Based Vehicle Optimization Using Machine Learning

Abstract

Traffic violations have caused immense harm. Moreover, numerous people, regardless of whether they are drivers or non-drivers have been affected. In a survey conducted by National Highway Traffic Safety Administration (NHTSA), ¾ of drivers admitted to over-speeding on all types of roads and were more prone to accidents. As a result, on maximum occasions, car accidents are usually the fault of rash driving. One of the choices to help activities that address these issues is to see how vehicle drivers perform when they are driving in terms of utilizing the vehicles resources. Utilizing car’s control unit data to analyze and measure driver’s performance is an issue that has acquired significance. Identifying automotive use profiles has become the focus of much research worldwide. This documentation depicts a machine learning classification model on vehicular information gathered from On Board Diagnostics-II device to recognize potential groups of automobile usage. Later with some refinement, the model introduced over 99% accuracy in distinguishing 3 user profiles (low, mid and high). This platform can acquire information from a vehicle and return its utilization profile, and stats on the driver’s usage. Which indirectly tells us whether someone driving is monetarily as well as environmentally economic or not.