Machine Learning for Predicting Prices and Empty Returns in Road Freight Transportation: Enhancing Efficiency and Sustainability

Road freight carriers constitute a crucial link within the economic chain, functioning as fully integrated players in supply chain management. Indeed, the proliferation of global exchanges, the emergence of new communication channels, and the advent of novel technologies have revolutionized its practices. Nevertheless, these carriers are susceptible to market fluctuations and uncertainties concerning prices and empty returns. Consequently, the prediction of these two indicators in order to proactively influence carrier profitability and sustainability remains a relatively underexplored topic. This paper bridges this gap by introducing a predictive analytical framework crafted to aid carriers in evaluating the profitability and sustainability of transportation requests. We employ two machine learning algorithms, namely artificial neural networks and XGBoost, utilizing road transport data from Morocco. Drawing from validation data, the developed framework demonstrates promising outcomes, providing managers with a systematic approach to analyzing business forecasts. The study also discusses the results and outlines potential directions for future research projects.