SNRLM: a cyber-physical based stepwise noise removal and learning model for automated quality assurance

SNRLM: a cyber-physical based stepwise noise removal and learning model for automated quality assurance

연구 분야: Verification

학회: International Journal of Information Technology

Businesses require processes that consistently meet high-quality standards to build a reputation for excellence. Modern technologies like IoT and cyber-physical systems are transforming product quality assurance, making it faster, more efficient, and adaptable to environments. This paper advocates for using cyber-physical systems (CPS) in production industries to enhance quality assurance. The reasons include ensuring production consistency, increasing efficiency, cutting costs by minimizing manual labor and inspections, improving defect detection accuracy, and offering scalability for evolving production demands. This paper, therefore, introduces a CPS system to assure product quality. For the experimental investigation, the paper used the casting product of defective and non-defective samples. A machine learning model is presented for quality assurance. The model is termed a stepwise noise removal and learning model (SNRLM) in which a digital filter is used to remove the noise from the captured product images. The result analysis was presented with a baseline model (SNRLM without digital filter) and achieved the highest accuracy of 97.8% for product quality assurance. The paper also presented a comparative analysis for state of art models and achieved a 5% improvement over existing models and concluded that the SNRLM may dramatically reduce inspection volumes, leading to economic gains.