AI-ML approach to uncover “hidden factory” of energy saving opportunity in semiconductor facility systems
Source abstract
Abstract
The semiconductor manufacturing industry is characterized by high energy consumption and stringent operational requirements, making energy efficiency a critical concern for both cost reduction and environmental sustainability. We explore how artificial intelligence (AI) and machine learning (ML) technologies can be leveraged to uncover hidden energy-saving opportunities and enhance the energy efficiency of facility systems in semiconductor plants. Through the integration of advanced data analytics, predictive modeling, and real-time optimization, AI/ML solutions can analyze complex operational datasets to identify inefficiencies, predict equipment performance, and suggest actionable improvements. Case studies and pilot implementations demonstrate how AI-driven insights enable smarter control of heat recovery as well as chiller and compressor systems, leading to significant reductions (6.5% and 10% validated annualized savings, respectively) in energy usage without compromising production quality or reliability. We also discuss challenges related to data quality, system integration, and model interpretability while highlighting the potential of AI/ML to drive sustainable innovation in semiconductor manufacturing facilities. The contribution of this work is an integrated, multi-level ML framework combined with expert validation, risk-governed deployment, and International Performance Measurement and Verification Protocol-aligned measurement and verification, enabling demonstrated energy savings across multiple fab utility systems.
Source: crossref · Quality: baseline
Research context
OpenAlex topics: Energy Efficiency and Management · Smart Grid Energy Management · Building Energy and Comfort Optimization
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
Affiliations
Micron Semiconductor Asia, Singapore
Schneider Electric SE, Singapore
References
Reference metadata was not supplied for this record.