Business Problem:

Client (Global Heavy Engineering OEM major) was facing inflated asset maintenance costs due to limited oversight into asset’s operation. They required a process for proactive monitoring to address the following business needs:

Mechanism for extended oversight and control through automation for predictive maintenance

No existing methods for data-driven decision making for proactive data analysis

The objective of this project was identified to make informed maintenance, component replacement, and repair related decision making.

Proposed Solution:

Cognilytic built predictive analytics-based condition monitoring solution that predicts the remaining useful life (RUL) of an equipment based on its usage. The following solution was offered:

Predict/ detect equipment abuse events (classification model) based on historic data abnormalities.

Survival Analysis using Weibull distribution.

Regression based remaining useful life (RUL) prediction of asset and end of life prediction.

Key Deliverables:

Our comprehensive ML models driven solution delivered the following results:

Inroads into situational intelligence by creating processes for anomaly detection, asset-based analysis, fleet-based analysis and site-based analysis.

Insights into operational intelligence through implementation of asset health score, repair time estimation mechanisms and failure prediction processes.

New actionable insights for improved business operations.

Basis this, the client realized the following business benefits:

Maximized utilization and life of existing and future equipment including increased uptime.

Lower operating costs by replacing planned maintenance and reactive maintenance with predictive maintenance.