IBM reliability leader highlights Internet of Things, data-driven decisions
Mary Bunzel, reliability champion for IBM Maximo, speaks at Fluke summit.
Major technological trends indicate a dramatic shift toward data-supported decisions on maintaining critical industrial assets, experts in reliability and thermography say. Such a shift will empower manufacturers to find hidden sources of revenue by avoiding unnecessary maintenance spending while at the same time maintaining quality.
One of the most important shifts is in growing ubiquity of sensors and wireless tools as well as the networking and storage infrastructure to communicate information about the health of equipment to decision makers.
"This capability to communicate with intelligence at the sensor level, at the fringe of the asset, is really changing the way that manufacturers are looking at their business streams," Mary Bunzel, reliability champion for IBM Maximo told attendees at the "Measurement of Innovation" Summit at Fluke Corp. headquarters in Everett Washington.
"This ability to hear what the equipment is saying and speak back to it is changing the way that we're interacting with our environment."
Acknowledging that a system that can trace, predict and prevent product quality issues through data capture and multiple adjustments is at present a goal, rather than reality, Bunzel said, however ample evidence exists that the so-called "Internet of Things" puts capturing and correlating data to support decisions within reach at many industrial plants.
"So in this case, smart manufacturing and the internet of things, the use of these sensors is enabling manufacturers to not do maintenance, which is - a lofty goal," Bunzel said. "I mean if you're not doing maintenance, you're not spending money, right? So we don't want to overmaintain."
"The Holy Grail, and the vision, the long-term view of this internet of things story is when we can look at product quality at the end and drive back a myriad of multiple adjustments to all these variables so that the product is in anticipated and manufactured on an automated basis used the learned intelligence of - current condition and its understanding of its impact on product. The whole idea is to identify quality issues - well in advance of - statistical models."
Bunzel used an example of a large automotive manufacturer who ran statistical models against data compiled from warrantee data. By running the analysis, a problem was discovered in certain parts that was recognized before the manufacturer even knew the issue existed.
The manufacturer was able to take proactive measures to correct the problem before it became critical because they were able to see the data.
"That's how compelling it is, business case-wise," Bunzel said. "So, you have predictive maintenance and quality capabilities that allow you to not only monitor the asset in real time but also to apply the right kind of maintenance at the right levels."
She said such system could be as simple as having maintenance software send an email when an anomaly occurs. The IBM Maximo asset management system has an email listener built in. At some point diagrams, images or other data could be attached.
"The whole idea is to get to the root cause of the failure more accurately so that you can make decisions about when to do the repair."
The recent "Measure of Innovation Summit" sought to engage the experts in dialogue about Fluke technologies and how technology can help industrial maintenance managers maintain uptime.
Unplanned downtime due to equipment failure can cost manufacturers up to 3 percent of their revenue, according the U.S. Federal Energy Management Program.
Manual methods of tracking equipment health to predict failures are time consuming and prone to errors and incomplete data, while existing computerized maintenance management systems can be costly and complex and often require significant IT resources to implement.
Fluke Connect® Assets changes the way equipment maintenance is documented, reported, and managed. The cloud-based wireless system of software and test tools that gives maintenance managers a comprehensive view of all critical equipment - including baseline, historical, and current test tool measurement data, current status, and past inspection data - enabling them to set up and sustain a predictive maintenance (PdM) or condition-based maintenance (CBM) system easily with minimal investment.
"Using Fluke-style preventive technologies [thermal imaging, vibration, etc.] should be about 25% of your maintenance effort, at a minimum, with two-thirds of the company's corrective efforts spent making fixes on what they find from those technologies [vs. on unplanned maintenance]," said Dr. Klaus M. Blache, Director of Reliability Center, University of Tennessee at Knoxville. "Most companies are below that. The opportunity for reliability, throughput, improved safety, better quality - it's an untapped resource, just doing that basic fundamental step out in the plant."