What question did this study set out to answer?

To develop and evaluate a machine learning framework for predictive maintenance in rotating equipment.

May 8, 2026Open Access

Machine Learning-Based Predictive Maintenance Framework for Rotating Equipment in Process Industries

View Full Paper

Ask AI

Bookmark

Authors

Randomized trial examines predictive maintenance in process industries, suggesting cost savings through advanced machine learning models.

To develop and evaluate a machine learning framework for predictive maintenance in rotating equipment.
Evaluated ML algorithms including Random Forest, Gradient Boosting Machines, SVM, LSTM, and a hybrid CNN-LSTM architecture.
Utilized triaxial accelerometers, thermocouples, and current sensors across three process plant sites in Northern India.
Conducted feature importance analysis using SHAP values and quantified maintenance savings through a cost-benefit model.
The CNN-LSTM hybrid achieved the highest macro-F1 score of 0.923 across seven fault classes, outperforming LSTM (0.891) and Random Forest (0.874).
Annual savings projected at ₹18.6 lakhs per monitored asset with a payback period of 14 months on investments.
Inference latency optimized at 47 ms per prediction cycle, meeting the real-time requirements of the SCADA system.

Dr. K. Sujatha (2026) studied this question.

View Full Paper

Ask AI

Bookmark

May 8, 2026Open Access

View Full Paper

Ask AI

Bookmark

Authors

Randomized trial examines predictive maintenance in process industries, suggesting cost savings through advanced machine learning models.

To develop and evaluate a machine learning framework for predictive maintenance in rotating equipment.
Evaluated ML algorithms including Random Forest, Gradient Boosting Machines, SVM, LSTM, and a hybrid CNN-LSTM architecture.
Utilized triaxial accelerometers, thermocouples, and current sensors across three process plant sites in Northern India.
Conducted feature importance analysis using SHAP values and quantified maintenance savings through a cost-benefit model.
The CNN-LSTM hybrid achieved the highest macro-F1 score of 0.923 across seven fault classes, outperforming LSTM (0.891) and Random Forest (0.874).
Annual savings projected at ₹18.6 lakhs per monitored asset with a payback period of 14 months on investments.
Inference latency optimized at 47 ms per prediction cycle, meeting the real-time requirements of the SCADA system.

Dr. K. Sujatha (2026) studied this question.

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

View Full Paper

Ask AI

Bookmark