There has been a rise in demand for affordable, sustainable last-mile delivery services in smart cities due to the upsurge in e-commerce and urbanization. Aside from many challenges to the traditional logistic infrastructure are environmental pollution, excessive operating costs, and traffic congestion in the roads. The research presented herein proposes an optimization model for artificial intelligence-enabled electric vehicles towards enhancing service reliability, reducing carbon emissions, and saving time on delivery. The EVRP solution model in accordance with city-specific parameters such as battery capacity, charging station availability, and prevailing traffic conditions has been developed by integrating Mixed Integer Linear Programming (MILP) and metaheuristic approaches, i.e., Genetic. The model performance was validated with actual delivery data sets on a simulated Lagos smart city network. Based on findings, relative to traditional routing methodologies, the AI-informed optimization model saved CO2 emissions by 31.4%, delivery time by 22.5%, and overall distance travelled by 17.8%. Additionally, the proposed system had flexibility with changing demand patterns and traffic flow, enhancing urban logistics resilience. These findings illustrate how electric vehicle routing using artificial intelligence can support policy initiatives in low-carbon smart cities, enhance business productivity, and promote environmental sustainability. Policymakers, logistics firms, and urban planners seeking to establish sustainable last-mile delivery networks in rapidly emerging cities will find this study helpful.
Popoola et al. (2025) studied this question.