Digital twin–driven optimization of manufacturing operations using real-time sensor data

Abstract

Abstract. With the advent of the digital twin technology, it is now possible to monitor and control manufacturing systems in an advanced way, as the technology has developed a virtual representation of physical processes in real-time. Nevertheless, the practical use of real-time sensor information in dynamic optimization has been an important challenge in traditional manufactures. This paper introduces a digital twin-inspired optimization model, which combines real-time sensor measurements with smart decisions to improve operational efficiency. According to the offered solution, there is constant alignment between the real and virtual systems through the IoT-enabled data acquisition. The digital twin has a predictive and optimization module that is used to analyze the behavior of the system, predict the performance, and adjust operational parameters on the fly. An optimization formulation to reduce production inefficiencies, energy use, and machine downtime and uphold system constraints is created. The model is tested with a representative manufacturing scenario that has sensor-driven inputs. Its results reveal that it performs better in terms of production performance, decreased downtime, and responsiveness compared with the traditional, non-dynamic optimization strategies. The proposed methodology provides a high-scaled and realistic approach to intelligent, data-driven manufacturing systems that align with the Industry 4.0 paradigms.