An Intelligent Cloud-Based IoT Framework Integrating Artificial Intelligence, Machine Learning, Computer Vision, Big Data Analytics, and Cybersecurity for Smart Industrial Management Applications

Abstract

Smart industrial management requires integrated systems that can monitor machine conditions, detect cyber threats, inspect product quality, and support operational decision-making. However, many existing Industrial IoT systems handle these functions separately, which limits coordinated industrial control and decision support. This study aimed to design and simulate a cloud-oriented Industrial IoT framework integrating machine learning, cybersecurity analytics, computer vision, big data-style preprocessing, and decision fusion for smart industrial management applications. The proposed framework used three data sources: a smart manufacturing dataset, an Industrial IoT network cybersecurity dataset, and a casting image dataset. Machine learning models were developed for machine condition prediction, cybersecurity models were used for binary intrusion detection and multi-class cyberattack classification, and a convolutional neural network was implemented for casting defect detection. The outputs of these modules were combined using a rule-based decision fusion engine to generate risk levels, alerts, decisions, and recommended actions. The results showed that Gradient Boosting achieved the best smart manufacturing performance with a weighted F1-score of 98.73%. The cybersecurity module achieved 99.71% weighted F1-score for binary intrusion detection and 90.62% weighted F1-score for multi-class cyberattack classification. The CNN model achieved 83.08% validation accuracy for casting defect detection. Overall, the proposed framework demonstrates the potential to support predictive maintenance, cyberattack detection, automated quality inspection, and simulated industrial decision support through an integrated Industrial IoT analytics workflow.