Agent-Based Simulation for Modeling Supply Chains: A Comparative Case Study

Keramydas Christos, Bechtsis Dimitrios, Aidonis Dimitrios

Contemporary Supply Chain (SC) networks are complex and dynamic environments comprising of stakeholders and their fragile relationships, along with the associated interconnecting products, cash and information flows. The main objective of the SC is the creation of added value at every single node of the network. In terms of coordination, taking into account the dynamic nature of SCs, central coordination techniques are considered of high risk, as they are not resistant to changes and have a single point of failure. Decentralized management is a precondition for creating flexible and effective networks, as nodes need to dynamically move in and out of the SC and every node has a decision-making ability. Decentralized networks have no single point of failure, as every node can support the decision-making process according to the actual state of the network. Agent-based (AB) simulation platforms create decentralized networks of peer nodes, provide native mechanisms of communication and coordination techniques and are suitable for simulating the complex nature of SCs.

In this paper, we first provide a literature review on AB simulation software used in supply chains modeling, we then describe an illustrative case study, and finally we develop two AB models using two distinct software platforms. The present work highlights the use of AB platforms agrifood supply chain modeling and can be used by academicians and practitioners as a practical example in order to incorporate AB techniques into the SC. The model-building details are provided for both platforms, as well as a critical analysis of the implemented models. The results demonstrate the incorporation of AB platforms to the SC ecosystem and compare the Agent-Based Modeling (ABM) capabilities of the alternative software platforms.