Hybrid Quantum–AI Optimization Models for Energy-Efficient and High-Yield Industrial Manufacturing Systems

Abstract: This study investigates hybrid Quantum–AI optimization models for improving energy efficiency and production yield in industrial manufacturing systems under realistic sensor-data constraints. The proposed framework couples (i) a physics-consistent energy model for electro-mechanical drives and process stages with (ii) an AI layer for state estimation and yield-risk scoring, and (iii) a quantum-ready optimizer for multi-objective setpoint selection. The decision problem is formulated as constrained multi-objective optimization that minimizes energy-per-part while maintaining yield, thermal safety margins, and throughput requirements. An empirical workflow is demonstrated using a cycle-level manufacturing sensor dataset containing speed, torque, power, thermal variables, vibration, tool-wear proxies, and pass/fail outcomes. Results show that a hybrid decision policy—representative of quantum-assisted combinatorial search over discrete setpoint candidates combined with AI-based quality-risk prediction—reduces energy per part while improving yield by stabilizing temperature and vibration regimes. The findings support an Industry 4.0 inference: hybrid Quantum–AI optimization can provide an auditable pathway for energy-aware control and high-yield production in high-mix manufacturing environments. Keywords: hybrid quantum–AI; multi-objective optimization; energy efficiency; yield; predictive quality; sensor data; Industry 4.0; manufacturing control.