AI-Driven Fleet Cost Forecasting: A Case Study in Logistics
Stouthuysen, Kristof Vanderstraeten, Bjarne Mellaerts, Mirthe
Stouthuysen, Kristof
Vanderstraeten, Bjarne
Mellaerts, Mirthe
Citations
Altmetric:
Publication Type
Editor
Supervisor
Publication Year
2025-08-18
Journal
Book
Publication Volume
Publication Issue
Publication Begin page
Publication End page
Publication Number of pages
8
Collections
Abstract
In early 2023, Maureen Philips, CEO of Actbel Logistics, faces ballooning cost volatility after pivoting from bulk chemicals to temperature‑controlled vaccine freight during COVID‑19. With 600 trucks, two distinct Business Units (Healthcare and Chemicals) and an impending ETS2 EUR0.15 /L carbon surcharge, the legacy unit-cost method is no longer capable of delivering accurate, forward-looking cost estimates. Maureen asks her Finance team to build a machine‑learning (ML) engine on three years of detailed truck‑month financials (≈ EUR185 million costs) and 1.2 million trip records to deliver granular, real‑time €/km forecasts. Students must (i) explore & combine the datasets provided, (ii) discover natural cost segments, (iii) benchmark supervised machine learning (regression) models, (iv) interpret feature importance with SHAP, and (v) translate analytics into contract‑pricing and fleet‑mix recommendations.
Research Projects
Organizational Units
Journal Issue
Keywords
Cost Forecasting, Logistics, Machine Learning