Cyber-physical systems (CPS) involve complex interdependencies between physical dynamics and computational control. Understanding and capturing these interactions is crucial for analyzing, predicting, and managing the performance of these systems, as it enhances understanding and facilitates optimization of operations. This study explores how to identify the buffer occupancy level of a cyber-physical serial production line, where buffer occupancy fluctuates over time. Although machine learning (ML) can learn temporal CPS behaviors, it typically requires large amounts of data and may struggle with highly nonlinear scenarios. To overcome these limitations, Quantum Machine Learning (QML) was explored and evaluated in comparison to ML models. Results indicate that the use of a hybrid QML model yields more accurate and efficient predictions of buffer occupancy levels compared to ML models, with a reduction in the mean squared error of 44.23% These indications explore new opportunities for QML in predictive modeling for CPSs and promote broader adoption of quantum computing
across various applications. They also demonstrate that adopting quantum technology in CPS enhances process control, improves production reliability, and achieves faster convergence with fewer parameters.

عنوان المؤتمر

International Conference on Sensors, Circuits & Instrumentation Systems

دولة المؤتمر

إيطاليا

تاريخ المؤتمر

30 مارس، 2026 - 2 إبريل، 2026

راعي المؤتمر

IEEE

معلومات إضافية

موقع المؤتمر الإلكتروني