Document Type
Research-Article
Journal Name
International Journal of Systems Science: Operations and Logistics
Keywords
Digital Logistics, Hidden Markov Models, Latent Dirichlet Allocation, patent analysis, technological evolution, topic modelling
Abstract
In the context of accelerating digital transformation and increasingly complex global supply chains, this study develops a multi-layered analytical framework to investigate the technological evolution and future trajectory of digital logistics. Using 107,202 patent records from the IncoPat database covering 1960–2023, the study integrates the entropy weight method, grey relational analysis, Latent Dirichlet Allocation (LDA), and a Hidden Markov Model (HMM) to identify core patents, extract technological topics, and forecast their dynamic evolution through 2028. The findings identify 27 major technological topics and reveal a significant transition from traditional operational optimization toward integrated intelligence and autonomous execution. Technologies related to optimization algorithms, intelligent logistics systems, automated handling, data networks, transport automation, and terminal communication emerge as the dominant drivers of future development. The results further indicate that digital logistics technologies are evolving from isolated functional applications toward deeply integrated intelligent systems characterized by automation, connectivity, and data-driven decision-making. Methodologically, the study contributes a dynamic and interpretable framework that combines topic modelling with probabilistic forecasting, extending existing patent-based technology evolution research beyond static analysis. Practically, the findings provide evidence-based guidance for technology investment, industrial upgrading, and policy formulation in the digital logistics sector. © 2026 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group.