Adaptability of Manufacturing Operations through Digital Twins

Manufacturing companies are under pressure to build faster and more efficient decision-making capabilities due to the rapidly changing customer demand and expectations. The conventional analytical models are no longer sufficient to capture the complexities of the supply chain. Companies are looking to embark in a digital transformation to address these challenges. One of the digital technologies that offer manufacturers a way to navigate this journey is digital twins, a virtual replica of an object, process or system. Our project focused on studying how digital twins can react to a complex and dynamic environment to create an adaptive mechanism and how can digital twins add value to increase operational efficiency. To answer these questions, we created a conceptual framework of digital twin, AI model and developed a learning feedback loop between simulation and artificial intelligence algorithm. We modeled the supply chain network by using data from a beverage industry and created what-if scenarios that involved varying customer demand and lead time through discrete-event simulation. The output of the simulation was fed into the AI algorithm. The AI prediction was simulated again and results were analyzed. Our research provides insights and discover value associated with adopting these technologies for better decision making. Our recommendation from this study will help supply chain managers understand that a digital twin and AI model framework can be developed, and can be utilized to foresee patterns in supply chain, and proactively take actions to resolve any bottlenecks and constraints.

Access full capstone paper on DSpace