Demand for regionally-produced food has seen tremendous growth in over the last decade, amid increasing consumer concerns over food safety, nutrition, origin, production practices and quality. Regional food systems provide economic support for small- and medium-sized farmers and help consumers become better-informed about their food, emphasizing the development of producer-consumer relationships and transparency with regard to production practices. In addition to these important social considerations, a sustainable and robust regional food system requires efficient and effective supply chain operations. However, most existing regional food supply chains (RFSCs) have not implemented appropriate supply chain management methodologies, and this has resulted in system-wide inefficiencies. Intermediated RFSCs, in which food is delivered to customers via a regional distributor, have recently become more prevalent. The role of the distributor, or “food hub”, is to provide a platform through which producers can efficiently and conveniently connect with customers. The food hub is also often responsible for ensuring transparency and facilitating information sharing and communication between producers and customers. Therefore, for a successful intermediated RFSC it is essential that the food hub manager selects his/her producers properly.

In this thesis, the impacts of variety of supplier selection policies on regional food system are discussed. We discuss what performance metric a RFSC should consider while evaluating the farmers. We also discuss the objectives of the food hub and the farmers. The food hub managers that we have interviewed have indicated that they do not have formal supplier selection policies. Instead, they randomly select suppliers that they believe will be able to fulfill their current demand. There is very little existing research on how to model the problem of supplier selection for regional food systems. In general, however, the literature suggests that multi-agent simulation (MAS) is a useful tool for studying supply networks and supply chain management methodologies. MAS is an approach to modeling and understanding complex systems that are composed of autonomous and interacting agents. Because a multi-echelon supply chain is a very complex social system, it is appropriate to use MAS to simulate supply chain behavior over time. Specifically, supply chains that are decentralized in command and control (such as RFSCs) are more appropriately captured using MAS techniques, rather than more traditional operations research methods. Many researchers have described supply chains by their constituent actors, activities, interdependencies, goals, and objectives, and they argue that systems possessing these components and structures are well-suited to analysis using MAS techniques. To study this system, we developed an agent-based model of a theoretical regional food system in NetLogo.