The growth of the productive sector has driven the evolution of supply chains into highly complex, dynamic and concurrent systems. Consequently, emerging structures have been developed through the integration of advances in information and hardware technologies with the aim of increasing effectiveness and efficiency, as Industry 4.0. In this context, although several works use a multi-agent approach, they offer limited information on the use of formal models to specification the system. Considering the potentially serious consequences of inaccurate integration specification, the use of formal tools to design the processes for these types of system is recommended. Considering supply chains as a system of discrete events, techniques derived from Petri Nets can be used for modelling purposes. Such techniques include Colored Petri Nets, which have been effectively tested in hierarchical modelling, analysis and control of distributed systems, characteristics suitable for the specification of a supply chain management system in an industry 4.0 scenario. In this context, this article presents a systematic procedure that integrates a group of guides, methods and techniques, to develop a representation of a system with problem solving abilities, thereby facilitating the construction of a supply chain management model with a multi-agent structure, which incorporates progressive refinements and permits specification of the desired level of detail. Finally, the advantages of the proposed model are presented via its application to a product assembly case study.