Optimizing energy and CO2 efficiency in last-mile delivery using hybrid fleet models

• Optimize CO 2 efficiency in last-mile logistics using prescriptive analytics. • Enhance supply chain resilience with NSGA-II multi-objective optimization. • Evaluate operational trade-offs in urban delivery via predictive modeling. • Compare RPASs and trucks to develop sustainable urban logistics strategies. • Provide actionable insights for low-carbon, data-driven transport planning. Effective urban delivery systems demand innovative approaches to reduce energy use and lower CO 2 . This study compares the environmental performance of hybrid and diesel trucks with quadcopter and fixed-wing remotely piloted aircraft systems (RPAS), employing a multi-objective optimization approach non-dominated sorting genetic algorithm II (NSGA-II) to identify optimal delivery routes balancing operational efficiency and sustainability. Given that existing solutions like e-bikes or electric vans may not be feasible everywhere, this research evaluates different vehicle types under various urban delivery scenarios. Using a synthetic dataset that simulates realistic conditions, the findings reveal that fixed-wing RPAS excel in long-range efficiency, while quadcopters perform better in short-range deliveries. Hybrid trucks are advantageous for larger loads, reducing emissions compared to diesel trucks. The results highlight key trade-offs in energy use and emissions, advocating for a mixed-fleet strategy tailored to specific logistics needs. This study provides actionable insights for sustainable urban freight planning and policymaking.