Designing a Digital Twin for a Mixed-Model Stochastic Assembly Line for the Reduction of Cycle Time

The Fourth Industrial Revolution has had a significant and far-reaching impact on the manufacturing industry. A substantial transformation has taken place within the manufacturing industry, with a notable shift from the conventional approach of mass production to a more bespoke model driven by the global market's demand for enhanced product diversity. This requires the redesign of assembly lines to enable the production of multiple product variants, thereby increasing their complexity. In order to effectively manage the increased complexity and avoid potential bottlenecks caused by longer cycle times, it is essential to implement a virtual system capable of real-time monitoring and fault detection. The current methods for reducing cycle time are deficient in their lack of utilization of real-time data inputs. This article presents a case study of a water bottling plant that employs a mixed-model stochastic assembly line. Two virtual systems, a digital shadow and a digital twin, were developed using MATLAB and SIMULINK as potential solutions. The two systems processed the identical input data in order to calculate cycle times. The results of the study indicate that the application of real-time data and digital twins can lead to a significant reduction in cycle times in a mixed-model assembly line, with an average improvement of 19% in comparison to the digital shadow.