Predict Equipment Failure With Advanced Analytics
Asset-intensive organizations are investing in advanced analytics to help them predict the failure of mission-critical equipment or assets.
Contributor: Susan Moore
Even in a digital world, physical infrastructure and equipment are still the backbone of many businesses. For asset-intensive industries such as agriculture, manufacturing, mining, oil and gas, transportation and utilities, the cost of unplanned downtime can be significant.
Predictive forecasting has long been the goal for those responsible for managing asset performance, but until recently, it has been plagued by technology limitations and project costs.
“Prediction is very difficult, especially about the future.” — Danish physicist Niels Bohr
According to Kristian Steenstrup, vice president and Gartner Fellow, advances in sensor technology, communications technology, information management and analytics are now making it possible to predict when a piece of mission-critical equipment will fail.
The result has been a surge of investment in predictive asset management solutions by small independent software vendors, such as Mtell, as well as large original equipment manufacturers, such as GE and ABB.
“It will continue to be a major focus for the next 10 years as globalization, regulatory oversight, social media scrutiny and complex value chains reduce the room for error in managing operational assets,” said Mr. Steenstrup.
How should organisations prepare?
Identify what data is necessary
Not all data is equally valuable when it comes to predicting the failure of assets, so it’s important to identify the data required for predictive asset management. Data used by advanced analytical engines to predict failure is largely data from operational technology (OT); that is, the time series production, equipment condition and event data that is used to control and monitor physical processes.
Predicting equipment failure does not always require data from the equipment itself, however. It’s sometimes possible to infer failure from seemingly extraneous data, such as production data, ambient temperature and data from peripheral equipment. Regardless, some form of OT data is required to effectively apply advanced analytics to predict equipment failure.
Invest in data science and analytics skills
Investing in data science and advanced analytics skills with a focus on predictive asset management will help support continuous improvement efforts for IT organisations, regardless of the status of current investments by the business. IT can then help ensure each project is successful and, more importantly, build the capability to deliver similar projects enterprise-wide.
Identify opportunities across the business
Most predictive asset management projects are initiated and executed at a local level. The end result may very well be a successful project, but it will be a successful local project, and the benefits will not extend throughout the enterprise. This is fundamentally the difference between how IT is managed (centrally) as opposed to OT (distributed).
IT should develop a strategy to identify existing projects and potential opportunities to deploy advanced analytics in support of predictive asset management, and offer to work with the business to achieve the desired results.
Given the advances in analytical tools to predict equipment failure, it’s inevitable that organizations with mission-critical assets will invest in advanced analytics to help ensure safe and reliable operations. IT should be proactive in identifying opportunities, building the expertise and developing a business plan.
Gartner clients can read more in the report ‘Using Advanced Analytics to Predict Equipment Failure.’
For more articles visit Smarter With Garnter website.