APM Digital Transformation

Asset Performance Management (APM) as we know today emerged as complementary solutions for CMMS, a tool to develop asset strategies based on their performances to improve the asset reliability and environment around it. In the past five years, APM has been associated with Digital Transformation (DT), Digital Twin (DTw), Industrial Internet of Things (IIOT) and Predictive Analytics (PrAx) as one of the key solution to improve the customer’s bottom line. Every software and consulting firm have their way of defining what digital transformation is and what role APM plays in it. Nevertheless, APM is now focused on Predictive Analytics with the sole purpose of predicting failure or provide a warning early enough for customers to consider a mitigation plan.

Many of the APM implementations never yield the desired benefit it is meant to deliver. One of the VP Engineering asked to review their APM Predictive Analytics Implementation, their problem was they would get few early warnings but in general, there was no reduction in the number of failures and the predictive analytics algorithm would still miss out on recognizing failure. We took this as a weekend project and evaluated the 44 Pump population which were connected to PrAx software. We also reviewed the ERP Data, work orders, and notifications for all 44 pumps and cataloged them for analytics.

For the 44 Pump Population, we observed,

§ 63 Evidences/ Symptoms

§ 24 Maintainable Items (Exl. Driver and Coupling)

§ 42 Primary Failure Modes (Most Occurring)

§ 226 Total Failure Modes

§ 318 Primary and Secondary Causes

Each of these elements was reviewed in detail and evaluated for possible detection method and their likelihood of being detected by sensors, human, and hydrodynamic calculations while the assets are in operation.

While Predictive Analytics provides relief in failure prediction one must bear in mind there are failure modes outside the horizons of sensor and human observations (While the asset is operational). These Failure Modes can be best managed thru a time-based checklist. The real challenge to the project was wrongly set expectations. Without any context to operating scenarios, the APM predictive Analytics solution will not yield benefits. As an owner-operator for a manufacturing facility one needs to balance out the expectation from an APM focused digital transformation.