International Journal Of Applied AI In Engineering Sciences • pp. 1-8
DEEP LEARNING FOR INTELLIGENT FAULT DETECTION IN INDUSTRIAL MANUFACTURING PROCESSES
Abstract
Deep learning has emerged as an important technology for intelligent fault detection in industrial manufacturing
processes. Modern manufacturing systems rely on complex machinery, automated production lines, and
interconnected sensors, making early fault identification essential for maintaining operational reliability and
product quality. Traditional fault detection techniques may depend heavily on manual inspection, predefined
thresholds, and expert knowledge, which can limit their effectiveness when industrial data are complex, noisy, or
rapidly changing. Deep learning techniques, including convolutional neural networks, recurrent neural networks,
autoencoders, and transformer-based models, can learn meaningful patterns from sensor measurements and identify
abnormal operating conditions. Previous research has demonstrated the potential of deep learning to improve fault
diagnosis, equipment monitoring, and industrial system reliability (Qiu et al., 2023; Zhao et al., 2024). This
research adopts a qualitative, literature-based approach to examine deep learning applications in manufacturing
fault detection. It focuses on fault identification, process monitoring, predictive maintenance, and the technical
challenges affecting real-world implementation. The findings indicate that deep learning can support automated
fault recognition and more timely maintenance decisions. However, limited fault data, class imbalance, changing
operating conditions, computational requirements, and model interpretability remain important concerns (Sahu et
al., 2024; Vashishtha et al., 2025). The study concludes that deep learning can strengthen intelligent manufacturing
when integrated with reliable sensor systems, appropriate evaluation methods, and established engineering
practices.
Record
- Identifier
- https://doi-services.com/1137/182
- Published
- 2026-11-03
- Version
- 1