Comparing Distance- and CO₂-Optimized Heuristics and Metaheuristics for the Multi-Depot Capacitated Vehicle Routing Problem in Agricultural Logistics
Comparing Distance- and CO₂-Optimized Heuristics and Metaheuristics for the Multi-Depot Capacitated Vehicle Routing Problem in Agricultural Logistics

Author(s): Giacinto Angelo SGARRO, Domenico Santoro, Francesco COLASANTO, Luca Grilli, Carlo Russo, Giulio Mario Cappelletti, Marsia CUSENZA, Antonio Vairo
Subject(s): Economy, National Economy, Business Economy / Management, Energy and Environmental Studies, ICT Information and Communications Technologies, Business Ethics, Green Transformation
Published by: Икономически университет - Варна
Keywords: capacitated vehicle routing problem; agricultural logistics; metaheuristic; ant colony optimization; heuristic; life cycle assessment; carbon footprint
Summary/Abstract: In this study, we tested three algorithms to solve a real-world agricultural logistics problem based on real geographic data. The task involves optimizing the collection and delivery of agricultural containers across multiple depots, using a fleet of capacity-constrained trucks. Two optimization scenarios were investigated. In the first, edge weights represented distances between points in kilometers. In the second, they represented emissions of CO₂ equivalent per route, calculated using Life Cycle Assessment (LCA) according to the Global Warming Total (fossil + biogenic) impact category of the Product Environmental Footprint (PEF), following the IPCC methodology. Each algorithm — a greedy heuristic (Eu) and two Ant Colony Optimization (ACO)-based metaheuristics (Meta 1 and Meta 2) — was executed under both distance-based and CO₂-based scenarios. The heuristic, being deterministic, was run once per matrix, while each metaheuristic was executed ten times per scenario. Total distance and CO₂ emissions were computed for each solution, and results were compared through boxplots and summary tables. Although differences between the two optimization criteria were minimal, CO₂-based optimization slightly outperformed distance-based optimization, yielding marginally lower total distances and emissions. Both ACO-based metaheuristic algorithms consistently outperformed the greedy heuristic, confirming their superior robustness. These findings suggest that, in this context, optimizing routes for CO₂ emissions or distance leads to comparable results, indicating that ecological and efficiency objectives may align in real-world logistics scenarios.

• Details
• Contents