The use of condition-monitoring data to identify when an asset is likely to fail (or when it doesn't need the maintenance work scheduled for it) necessitates a significant financial commitment as well as a mindset shift. But the ability to do advanced analytics aimed at eliminating unscheduled downtime soon won't be a nice-to-have for industrial production facilities; it will be a need-to-have. That's according to the four industry panelists featured in the session "No Downtime: The Power of Predictive Analytics” at the Smart Industry 2015 Conference in Chicago.
"If you don't get in on this movement, you will be left behind," said David Bartlett, CTO of GE Aviation. Bartlett noted that GE CEO Jeff Immelt likes to comment that the company "went to bed as an industrial company and woke up as a software and analytics company." For OEMs, the ability to track how well equipment is performing in the field and troubleshoot problems in real time for customers will be a competitive necessity. And for all manufacturers, being able to track where crucial assets (equipment as well as personnel) are and their operating status will enable bottom-line-changing efficiency gains.
Six weeks’ notice
Sarah Prinster, vice president of marketing at Apprion, developer of the ION system for managing industrial wireless infrastructure, noted that her company's technology can help manufacturers identify which machines are on track to fail prematurely as early as six weeks out from a predicted failure. That kind of information can make for a highly compelling argument for investing in condition monitoring and analysis. "Our customers are really realizing the benefits of predictive analytics," she said.
Chris Witte, a senior vice president at chemicals producer BASF Corp., said that department heads looking to "sell" management on investments related to the Industrial Internet of Things (IIoT)—including condition-monitoring sensors and geolocation technology—can build their case for initial outlays by emphasizing the safety and security benefits. Expanded use of sensors on assets large and small can mean better tracking of real-time field conditions, potentially keeping personnel out of harm's way and allowing for real-time corrective actions that can help companies avoid unplanned downtime and ensure regulatory compliance.
"Look for incidents that took you down, that you weren't ready for, and justify investments based on preventing them," Witte advised. "Show how you can keep it from happening again."
Start small, scale up
That was the approach taken at BASF, Witte said: Start small and scale up. Adopting a long-term, strategic approach to building a smarter and more connected manufacturing environment, the company has come to view analytics technology as a vital tool for everything from monitoring construction work to aiding in knowledge capture and transfer and even recruiting new talent.
For companies farther along on their IoT journey—those that have literally and figuratively bought into condition monitoring and associated analytics—one of the top challenges is managing the volume of data that all of their connected assets generate.
"If you look at a jet engine," Bartlett commented, "you can produce half a terabyte of data on one trans-Atlantic flight." The imperative for businesses is to look for anomalies—the outliers—rather than capture and store everything indefinitely, he suggested. "That's something you have to face as a company if you're seriously moving into this space," he said.
"What's the minimum amount of data you need to effect change?" he asked. "The art of Big Data lies in finding the least amount of data you need to do so.”