• Hritik Thorat
• Dr. D.T Bornare

Abstract :The development of smart manufacturing and Industry 4.0 has emphasized the utilization of intelligent manufacturing tools, techniques, and methods such as Predictive Maintenance (PM). The predictive maintenance function facilitates the early detection of failure and errors in machinery before they reach critical phases. Proper maintenance keeps the life cycle cost down and guarantees proper operations and good order internal logistics. This dissertation proposes a data-driven predictive maintenance framework tailored for IoT-enabled smart companies. Unlike existing research that focuses on anomaly detection through simulations, our study presents an end-to-end PM framework evaluated with real-world manufacturing data. The framework comprises five layers: Data Preprocessing and Feature Extraction, Feature Ranking and Selection, Clustering for Anomaly Pattern Mining, Supervised Anomaly Detection, and Anomaly-triggered Remaining Useful Life (RUL) Prediction. Combining unsupervised clustering with supervised deep learning, the framework effectively identifies anomalies and estimates equipment RUL. Testing revealed significant accuracy in tool wear and RUL prediction using LSTM variants, demonstrating the framework's effectiveness across various use cases.

Predictive Maintenance, Anomaly-Onset Aware RUL Estimation, LSTM Variants, Data driven Model

"PREDICTIVE MAINTENANCE DEEP LEARNING FRAMEWORK FOR IOT-ENABLED FOOD PROCESSING EQUIPMENT", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.11, Issue 6, page no.e88-e102, June-2024, Available :http://www.jetir.org/papers/JETIR2406415.pdf

"PREDICTIVE MAINTENANCE DEEP LEARNING FRAMEWORK FOR IOT-ENABLED FOOD PROCESSING EQUIPMENT", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.11, Issue 6, page no. ppe88-e102, June-2024, Available at : http://www.jetir.org/papers/JETIR2406415.pdf