Work

A simulation-driven approach to last-mile logistics analysis considering microlevel operational sequences for efficient delivery planning

Year: 2026
Type: article
Abstract: Supply chain management is becoming more challenging due to shifting client expectations and ongoing technological advancements. Failure to implement these changes can lead to resource mismanagement, delivery delays, and high operational costs. To address these challenges, this study presents a hybrid simulation framework combining Discrete Event Simulation (DES) and Agent-Based Modeling (ABM) for efficient decision-making. The proposed framework uses micro-level task modeling with the Maynard Operation Sequence Technique (MOST) to estimate delivery service times. In addition, Monte Carlo simulation examines output variability, and the Wasserstein distance-based double bootstrap method provides statistical validation. This holistic approach enables informed, data-driven decisions to avoid resource underutilization, ensure proper staff allocation, and maintain service consistency. Simulation scenarios revealed that incorrect planning can lead to unmet delivery commitments or excessive staffing costs. Experimental results indicate that the micro-level framework outperformed the existing simulation (Macro/Traditional) in operational efficiency. The proposed framework is benchmarked against exact MILP-VRP (Mixed Integer Linear Programming) optimization and heuristic ALNS (Adaptive Large Neighborhood Search) approaches. Compared to existing approaches, the proposed approach achieved higher efficiency by increasing deliveries by 28% with fewer fleets, increasing vehicle utilization to 98%, reducing elapsed time, and reducing delivery penalties. This approach was thoroughly validated using GIS-based data, experimentation, and performance analysis, showcasing its effectiveness in optimizing operational efficiency and profitability.
Language: English
Subfield: Transportation
Open Access status: hybrid
Estimated APC paid: $3,090
OpenAlex ID: W7225002467
FWCI:
Cites: 47
Cited by:
Related to: 10