Predictive Analytics Success in Power Generation
By Stanley Moses, December 16, 2015
Last month, @Bahwan CyberTek (BCT) successfully completed a pilot using BCT’s predictive analytics platform – RETINA with a major power generation company. The pilot was to demonstrate the capability of predictive analytics in power generation industry.
This is a good example for an IT-OT integration – though we just scratched the surface.
The goal was to generate new insights for operations & maintenance teams such as detecting performance degradation, predicting failures, identifying root causes, mapping economic analysis, and prognostic duration to failure. Years of asset data from Historian, EAM & Mobile applications were integrated with RETINA to present a 360 deg view for the asset (condenser).
Predictive tools helped convert data into information helpful for real-time decision making – due date for maintenance, operator actions, schedule planning. The solution highlights abnormal conditions days/hours (depending on the issue identified) before the failure event. As a result of the pilot, the unit is better prepared on most situations than fire-fighting after failures. The power generation company is now evaluating to expand similar approach for all their assets.
How is this IT-OT? – Data from an operational application such as control systems (stored in historian here) is used in conjunction with a transactional application such as Enterprise Asset Management (EAM) to deliver value for different stakeholders in the power generation unit. Which otherwise is seen as silos and not contextualized. Surely there are lot more ideas to utilize ERP data (purchase, vendor, HR) for improved outcomes. Any thoughts from you?