Machine Learning-Based Data Strategies in Automotive After-Sales Services: Systematic Literature Review

Abstract

This study presents a systematic review on machine learning-based data strategies available in the automobile after-sales service research area to offer insights about currently published studies, their predominant trends, related challenges and their future directions. Using systematic literature review (SLR) methodology, 508 articles were initially sourced from Scopus, with 23 articles selected as meeting the inclusion criteria for deeper analysis. This paper aims to look into machine learning making methods and tools, datasets and their applicability in fostering data strategies for after sales service. Results show that supervised learning techniques such as Support Vector Machines, Random Forests and Neural Networks are especially useful for predictive maintenance and demand forecasting. On the other hand, it is utilized on data without labeled output by unsupervised and reinforcement learning for anomaly detection and decision making too. This also contributes towards enhanced efficiency, better utilization of resources, and higher overall customer satisfaction in automotive after-sales services through deep learning via TensorFlow, Keras, and MATLAB, along with diverse data sets. The elevated application of the ways to manipulate data aspire for a better approach to achieve increasing operational efficiency, improved customer satisfaction and agile solutions in real-time. Variety of data, privacy challenges and a lack of better standards, however, are limiting wider usage. Further studies in particular concerning real-time data and holistic studies, especially covering dealers as representative of after-sales service, shall be at the forefront of future studies. This study bridges the gap by synthesizing the machine learning applications in automotive after-sales services to establishing valuable insights that will guide academia and praxis in applying more practices and methods based on data and prediction of customers to boost the sustainability of after-sales services in the automotive sector.