An agent-based model of regional food supply chain disintermediation

Regional food hubs provide logistics services for small and mid-sized food producers, giving them the ability to reach larger markets and customers than they can reach on their own. However, once food hub managers have helped to establish connections between producers and new customers, they often find themselves cut out of the regional food supply chain when the farmers decide to sell their products directly to the customers, thereby avoiding the food hub’s service fees. Widespread disintermediation can eventually lead to food hub failure, which can disrupt the entire regional food system. This paper describes an agentbased model that incorporates reinforcement learning to study disintermediation behavior in a regional food supply network in Iowa. The model is designed to serve as a decision support tool for food hub managers, allowing them to simulate the effects of various supply chain management strategies on producer decision making and long-term system success.