A survey of machine learning applications in advanced transportation systems: Trends, techniques, and future directions

Abstract

In recent years, artificial intelligence (AI) has revolutionized numerous sectors, including advanced transportation systems (ATS). This paper presents a comprehensive review of the latest machine learning (ML) applications within ATS, encompassing air, marine, and land transport modes. The review systematically categorizes and evaluates ML applications in four key subdomains: more-electric aircraft (MEA), all-electric ships (AES), high-speed rail (HSR), and electric vehicles (EV). A total of 124 articles were analyzed, spanning January 2014 to December 2023, to identify the global focus and results of ML in ATS. Our findings reveal that ML methods significantly improve predictive maintenance, energy management, fault diagnosis, and system optimization in ATS. However, the adoption and integration of ML face challenges related to data quality, model complexity, and real-time implementation. This review serves as a multidisciplinary research roadmap, considering ATS as a whole and taking a broad perspective of ML applications in ATS; highlighting open challenges and future directions, including dealing with data limitations, computational demands, applying transformers for time series forecasting, applying other emerging ML methods in ATS, and combining different ML approaches. The insights provided aim to facilitate further adoption of ML by both academia and industry, ultimately contributing to the evolution of intelligent and efficient transportation systems.