Artificial intelligence and digital technologies in railway systems: a systematic literature review using the TCCM framework

Purpose Artificial intelligence (AI) has transformed industries worldwide, driving digital innovation operational efficiency, safety, reliability and overall sustainability. Additionally, associated technologies such as machine learning, deep learning, digital twins and the Internet of Things (IoT) have further accelerated the digital transformation across railway systems. The literature in this domain is still at a developmental stage and remains fragmented across multiple application domains. This limits the understanding of theoretical foundations, context and methodological approaches. Design/methodology/approach To address this gap, this study conducts a hybrid systematic literature review using thematic analysis of bibliometric data from 121 Scopus records followed by application of the Theory–Context–Characteristics–Methodology (TCCM) framework on a subset of the literature identified through systematic screening based on Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. Findings The theoretical and conceptual development in this domain is strongly focused on the themes relating to predictive maintenance, infrastructure monitoring and operational optimisation. Novel technologies such as machine learning, deep learning, digital twins and the IoT emerge as the most frequently discussed technologies in the context of adoption within existing systems. The integrated TCCM analysis identifies three broad trajectories: a technology-integration trajectory centred on digital twins, cyber-physical systems and Industry 4.0/5.0; an algorithmic trajectory centred on AI, machine learning, deep learning and prognostics; and a decision-optimisation trajectory involving multi-criteria decision-making, fuzzy logic and intelligent transport systems. There is a comparatively higher concentration of literature on developed economies. Freight and logistics, despite emerging as an important thematic area, remain theoretically and methodologically fragmented, while emerging economies and several passenger- and sustainability-oriented applications receive limited attention. Methodologically, the literature is dominated by reviews, conceptual frameworks, modelling, simulation and experimental studies, with limited longitudinal evidence on the organisational and strategic consequences of AI adoption. There remains scope for stronger theoretical grounding, focusing on freight logistics, emerging technologies and longitudinal investigations to advance the literature. Originality/value This study provides a structured synthesis of the fragmented literature on AI and digital technologies in railway systems through the TCCM framework. It identifies key research gaps and proposes a future research agenda emphasising stronger theoretical foundations, freight-oriented applications, emerging technologies and longitudinal empirical investigations to support the development of intelligent and sustainable railway ecosystems.