• Tarika
• Dr. Sonu Agrawal
• Prof Shankar Saran Tripathi

The increasing deployment of Industrial Internet of Things (IIoT) devices has resulted in the generation of massive volumes of sensor data, offering significant potential for predictive maintenance and fault detection. This research aims to develop machine learning models that can effectively anticipate equipment failures and detect anomalies using sensor data from IoT devices. The study involves collecting and preprocessing sensor data from secondary sources, analyzing this data to detect patterns and trends, and creating machine learning models for predicting equipment faults. To enhance prediction accuracy, various machine learning techniques were employed. Analytical techniques such as histogram plots and six-sigma analysis were used to understand data distribution and identify potential outliers. Feature engineering techniques, including extracting timestamp, sensor ID, and value, were used to improve the models' capabilities. Grid search CV was applied for hyperparameter tuning, particularly for Support Vector Machine (SVM) models, to optimize model performance. The research explored both unsupervised and supervised learning algorithms. Unsupervised learning methods, such as Isolation Forest and K-Means Clustering, were effective in detecting anomalies. Supervised learning algorithms, including Support Vector Classifier and Decision Tree Classifiers, were used to evaluate fault detection accuracy. Performance metrics like accuracy, precision, recall, and F1-scores were employed to assess model performance. The Decision Tree Classifier emerged as the most accurate model for supervised learning with an accuracy of 0.5, while the Isolation Forest was the best-performing model for unsupervised learning in detecting faults. These findings demonstrate the importance of selecting appropriate machine learning models and techniques for effective fault detection in industrial IoT systems. The study underscores the potential of machine learning to enhance predictive maintenance and operational reliability in industrial settings, thereby improving uptime, reducing costs, and enhancing worker safety.

Industrial Internet of Things (IIoT), Machine Learning, Predictive Maintenance

"Predicting Equipment Failures and Detecting Anomalies in Industrial IoT Systems with Machine Learning", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.11, Issue 7, page no.a494-a499, July-2024, Available :http://www.jetir.org/papers/JETIR2407047.pdf

"Predicting Equipment Failures and Detecting Anomalies in Industrial IoT Systems with Machine Learning", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.11, Issue 7, page no. ppa494-a499, July-2024, Available at : http://www.jetir.org/papers/JETIR2407047.pdf