Abstract

Purpose: Rapid growth of e-commerce in India has intensified urban congestion and air pollution, positioning last-mile distribution as a critical area for decarbonisation. This study develops a low-emission last-mile distribution network design framework tailored to Indian urban contexts, where conventional designs have primarily prioritised cost and service speed. Research design, data, and methodology: A multi-objective optimisation framework is proposed using a mixed-integer programming model that simultaneously minimises total distribution cost and CO₂ emissions while maintaining predefined service levels. The model captures Indian-specific characteristics, including heterogeneous vehicle fleets (conventional vans, two-wheelers, and electric vehicles), micro-fulfilment centres, congestion-induced time-dependent travel speeds, and regulatory restrictions on vehicle access in dense urban areas. A hybrid solution approach combining ε-constraint scalarisation with a tailored metaheuristic is employed to approximate the Pareto frontier for realistic problem sizes. Results: Numerical experiments based on representative data from major Indian metropolitan areas show that low-emission network configurations can reduce last-mile CO₂ emissions by 18–30% compared to current practices, with cost increases typically remaining below 7% when service levels are preserved. Conclusions: The study provides a decision-support framework for e-commerce firms and urban policymakers to evaluate context-specific trade-offs between cost, service quality, and environmental performance in Indian last-mile logistics.