AN ANALYTICAL APPROACH-BASED DECISION SUPPORT FRAMEWORK FOR BUILDING SECURE FOOD SYSTEMS

Abstract

Food security refers to a system's capability to provide all individuals with social, physical, and economic access to sufficient, safe, and nutritious food that satisfies their needs and preferences for a healthy and happy life. Food security encompasses various dimensions, including accessibility, availability, utilization, and stability of food sources. Recent growth and diversity of food systems disruptions necessitate an efficient decision-making process to strengthen these systems' ability to withstand their consequences. The interconnection between the food systems' efficiency, robustness, and resilience profoundly influences the food system's security. This intricate and diverse nature of the food systems increases the decision-making process's complexity and underscores the necessity for integrated approaches to assess and enhance food systems' performance. In recent years, there has been a growing recognition in today's business of the importance of food systems assessment in making informative decisions to mitigate the risk of disruptions and enhance food security. Several methods and approaches have been developed for food security improvement. This study aims to advance the state-of-the-art in food security assessment and improvement by presenting three research contributions that are anticipated to provide food systems management and stakeholders with integrated approaches for making informed decisions considering the multidimensional structure of the food systems. The first research contribution focuses on the relationship between food system efficiency and resilience and their significance for food security and proposes an integrated framework for simultaneously enhancing these food security measures. The framework integrates penalized regression with data envelope analysis (DEA) to evaluate and improve food system efficiency and promote resilience. The mathematical and operational performance of the proposed framework was evaluated using the global agricultural food systems. The integration of these methods has shown an appropriate level of applicability and capability to overcome major drawbacks of existing frameworks that might affect their applicability on various scales of food systems. The second research contribution focuses on the interrelationship between the efficiency and robustness of food systems and proposes an integrated approach for providing stakeholders with insightful data for balancing this relationship and promoting food security. This relationship is inverse by nature, meaning that over-optimization of resources might impact resource availability for developing contingency plans. On the other hand, under-optimization might impact the food availability required to ensure appropriate food security levels. The mathematical and operational performance of the proposed framework was evaluated using the global agricultural food systems. The approach has shown notable effectiveness in identifying inefficient robustness dimensions, providing management and decision-makers to investigate several opportunities for improving robustness and promoting food security. The third research contrition focuses on the impact of the high-dimensional nature of food systems assessment and provides a dimension reduction-based approach for overcoming the "curse of dimensionality" and providing a reliable assessment of food security. The proposed approach utilizes the two clustering approaches to enhance the usability of the food security assessment in providing decision-makers and stakeholders with opportunities to identify the best practices in food security enhancement. The proposed approach has been evaluated using the same Global Food Security Index (GFSI) data set for a fair comparison. The approach has shown adequate performance compared with the GFSI in providing a comprehensive measure that considers the output and characteristics of the food systems. The three approaches have shown reasonable flexibility and scalability features compared to some existing approaches. These features help ensure these approaches' broad applicability across several operational levels in other industrial and service systems. Several conclusions, recommendations, and future research have been drawn and reported.