An Intelligent Model for Fleet Management by Use of Sensor Enabled Tags Integrated With GPRS Technology

Abstract

Fleet management is used to optimize transport functions and is implemented using the integration of technologies. This study seeks to address the issue of fuel adulteration. The high influx of vehicles has resulted in high consumption of fuel where fraudulent dealers exploit consumers by selling adulterated fuel. As a result; adulterated fuel leads to rapid wear and tear of parts, engine failures and complete breakdown leading to high cost of maintenance through frequent repairs and replacement of vehicle engines. Other consequences of fuel adulteration include the release of toxic gases into the air which are hazardous to the ecosystem and also a health hazard and loss of revenue for the government. Methods used previously to check for fuel  adulteration include chemical methods; laboratory-based methods that examine chemical properties of the fuel, mechanical methods such as the use of sensors which involve installation of testing gadgets and mobile testing kits that can be used during fuelling. Presently the world is moving towards knowledge-based applications where there is a huge consumption of data; data from different sources make up big data that is analyzed to identify trends and make predictions. This study proposes the use of supervised machine learning that analyses vehicle data and is able to predict the number of vehicles using adulterated fuel based on the performance of a vehicle that is measured parameters by some onboard parameters. An experiment was performed using three classifiers and Random Forest classifier gave the best results examined by confusion matrix.