Designing cost-effective and environmentally responsible saffron supply chains: A two-stage data-driven approach for public and service sector decision-making

Modern agricultural supply chains face increasing complexity due to escalating environmental pressures and growing demand for affordable, high-quality food products. These challenges underscore the critical need for strategic, quantitative decision-making frameworks that enhance operational efficiency and sustainability, particularly within socio-economic sectors. Saffron, as a unique agricultural product with specific biological and production characteristics, requires a specialized and well-structured supply chain network to address these demands. This study develops an advanced data-driven decision-support framework designed to assist managers in public and service sectors in effectively designing and managing saffron supply chains that balance efficiency with environmental responsibility. This framework uniquely integrates supplier evaluation, environmental impact, demand prediction, and network design, including spoilage considerations, addressing gaps in the literature. The framework consists of two integrated phases. The first phase employs a hybrid supplier evaluation model that combines data envelopment analysis with autoencoder-based neural networks to assess supplier performance in terms of operational efficiency and environmental stewardship, even when faced with incomplete or uncertain information. In the second phase, selected suppliers are incorporated into a multi-objective optimization model that guides strategic network design by simultaneously minimizing costs and environmental impacts. A life cycle assessment using the ReCiPe methodology evaluates environmental effects comprehensively across the supply chain. Additionally, long short-term memory neural networks provide accurate demand forecasting to align supply strategies with market fluctuations. The results indicate that the model identifies efficient facility locations, demonstrates how modest cost adjustments can enhance environmental performance, and supports managerial decision-making by balancing cost efficiency with sustainability objectives.