A modular hybrid simulation framework for complex manufacturing system design

Abstract

For complex manufacturing systems, the current hybrid Agent-Based Modelling and Discrete Event Simulation (ABM–DES) frameworks are limited to component and system levels of representation and present a degree of static complexity to study optimal resource planning. To address these limitations, a modular hybrid simulation framework for complex manufacturing system design is presented. A manufacturing system with highly regulated and manual handling processes, composed of multiple repeating modules, is considered. In this framework, the concept of modular hybrid ABM–DES technique is introduced to demonstrate a novel simulation method using a dynamic system of parallel multi-agent discrete events. In this context, to create a modular model, the stochastic finite dynamical system is extended to allow the description of discrete event states inside the agent for manufacturing repeating modules (meso level). Moreover, dynamic complexity regarding uncertain processing time and resources is considered. This framework guides the user step-by-step through the system design and modular hybrid model. A real case study in the cell and gene therapy industry is conducted to test the validity of the framework. The simulation results are compared against the data from the studied case; excellent agreement with 1.038% error margin is found in terms of the company performance. The optimal resource planning and the uncertainty of the processing time for manufacturing phases (exo level), in the presence of dynamic complexity is calculated.