Collaborative, Distributed Simulations of Agri-Food Supply Chains. Analysis on How Linking Theory and Practice by Using Multi-agent Structures

Simulations help to understand and predict the behaviour of complex phenomena’s, likewise distributed socio-technical systems or how stakeholders interacts in complex domains. Such domains are normally based on networked based interaction, where information, product and decision flows comes in to play, especially under the well-known supply chains structures. Although tools exist to simulate supply chains, they do not adequately support multiple stakeholders to collaboratively create and explore a variety of decision-making scenarios. Hence, in order to provide a preliminary understanding on how these interaction affects stakeholders decision-making, this research presents an study, analysis and proposal development of robust platform to collaboratively build and simulate communication among supply chain. Since realistic supply chain behaviours are complex, a multi-agent approach was selected in order to represent such complexities in a standardised manner. The platform provides agent behaviours for common agent patterns. It provides extension hotspots to implement more specific agent behaviour for expert users (that requires programming). Therefore, as key contribution, technical aspects of the platform are presented, and also the role of multi-level supply chain scenario simulation is discussed and analysed, especially under de context of digital supply chain transformation in the agri-food context. Finally, we discuss lessons learned from early tests with the reference implementation of the platform.