Towards sustainable logistics in Turkey: A bi-objective approach to green intermodal freight transportation enhanced by machine learning

Abstract

Transportation is a critical contributor to carbon emissions, with road transportation playing a dominant role due to its dense network and versatility. However, the overreliance on road transportation has led to congestion, impacting reliability. As international trade grows, the demand for sustainable logistics practices intensifies. Intermodal transportation systems have emerged as a promising solution, harnessing different modes to reduce emissions and environmental impact while optimizing costs. It is important to underscore the significance of mode combinations in achieving environmental goals, aligning with the broader concept of environmental sustainability that encompasses economic and social dimensions. This article contributes to this evolving landscape by presenting a bi-objective intermodal transportation problem focusing on carbon emission reduction. Leveraging machine learning algorithms, including multiple linear regression, support vector regression, decision tree, and random forest, we predict transportation-based CO₂ emissions, offering environmentally friendly logistics plans. Our research responds to the call for green intermodal transportation, addresses financial incentives, emphasizes profit maximization, and reflects the growing influence of government policies. This paper outlines our methodology, presents a real-world case study, and offers computational results, underscoring the significance of sustainable intermodal transportation in the context of global climate goals and government initiatives.