An Empirical Study of Car Failure Causes Based on Machine Learning Methods

Abstract

Car breakdowns are a widespread issue for car owners, circulation roads and maintenance companies in all the world, leading to significant economic and safety implications. In order to develop more effective preventive maintenance strategies and minimize these consequences, this study has aimed to gain valuable insights into the primary causes of car breakdowns. In order to achieve this, a survey targeting drivers from all Moroccan cities as a sample is conducted, focusing on breakdown types, their frequencies, as well as maintenance practices. The study has employed significant variables, including driver demographics, car brand and model, horsepower, fuel type, and maintenance practices, in a prediction test to anticipate car breakdowns. Statistical and machine learning algorithms, such as regression and classification analysis, have been used to identify six underlying factors contributing to car breakdowns in Morocco. This research provides crucial insights for car owners, maintenance companies, and policymakers to improve preventive maintenance strategies and decrease the economic and safety impacts of car breakdowns that could be applied in all the world.