Edge AI-Based Smart Factory Development for Carbon Emission Reduction

Abstrak

Manufacturing industries are facing increasing pressure to reduce carbon emissions while maintaining high levels of productivity and operational efficiency. In response to these challenges, Edge Artificial Intelligence (Edge AI) has emerged as a promising technology for enabling real-time analytics and intelligent decision-making within Smart Factory environments. This study aims to develop an Edge AI-based Smart Factory framework for monitoring, optimizing, and reducing industrial carbon emissions through intelligent energy management. The proposed framework integrates Industrial Internet of Things (IIoT) sensors, edge computing devices, artificial intelligence algorithms, and carbon monitoring modules to collect, process, and analyze manufacturing data in real time. Machine learning models, including Random Forest, XGBoost, and Long Short-Term Memory (LSTM), are deployed on edge devices to predict energy demand, detect operational inefficiencies, and optimize production activities. The framework is evaluated using energy efficiency, carbon reduction, operational performance, and AI model accuracy metrics. Experimental results demonstrate that the proposed system significantly improves operational efficiency, reducing energy consumption from 1000 kWh to 820 kWh and decreasing machine idle time from 18% to 7%. Furthermore, carbon emissions are reduced from 700 kg/day to 540 kg/day, representing a reduction of 22.9% compared to conventional factory operations. The LSTM model achieved the highest predictive accuracy of 95%, supporting effective real-time optimization and decision-making. These findings indicate that Edge AI can effectively support sustainable manufacturing by enabling intelligent energy management, real-time operational optimization, and carbon-aware production decisions, thereby contributing to the development of greener, more efficient, and more resilient Smart Factory ecosystems.