BETA
This is a BETA experience. You may opt-out by clicking here
Edit Story

Real-time Repairs: Why The Internet Of Things Is A Game-Changer For Manufacturers

Pega

By Nick Clunn

Knowing whom to call when equipment breaks is important. Having that person call you to schedule a tune-up before the equipment goes down is even better.

For all of the virtues of preventive maintenance, there’s one major drawback. The traditional approach to maintenance relies on accepted practice—the oil change every 3,000 miles, for example—rather than on actual conditions.

But that’s beginning to change, thanks to the combination of big data and the Internet of Things (IoT). Manufacturers armed with Internet-connected sensors and data-crunching software can now watch what happens inside the products they make after they leave the factory, and predict what could happen next.

With such insight, manufacturers can diagnose problems remotely, service products only when needed and deliver better customer service.

High Stakes

GE has invested heavily in systems that can deliver such outcomes for its large industrial clients—companies that generate electricity, run trains and fly airplanes.

To avoid unexpected breakdowns and unnecessary maintenance, GE creates virtual models of the equipment being monitored—such as an aircraft engine or a gas turbine—with technology called Digital Twin. In addition to helping engineers assess present conditions (whether or not a product requires maintenance, for example), Digital Twin provides suggestions to make operations more efficient, based on data delivered by sensors that is continually analyzed.

Applied to a digital wind farm designed by GE, Digital Twin uses data collected by embedded sensors to develop maintenance schedules that are unique to each wind turbine. It also prolongs the life of each turbine by maintaining ideal conditions.

The Long View

Beyond basic maintenance, the latest IoT software is powerful enough to analyze “event streams” across entire fleets or product lines to anticipate problems and take action before they occur. In other words, what’s happening with the whole is informing what will likely happen to the individual.

“If I can attach myself to the event stream and respond to patterns in the event stream—that’s how I prepare myself for a potential recall or part change. Or, that’s how I notice a trend and get ahead of it,” said Don Schuerman, chief technology officer at Pegasystems, a company that helps global enterprises in manufacturing and other industries enhance the customer experience.

How to respond to that trend—and manage the complex logistics involved—is also a task for intelligent software, Schuerman said. A broken part may seem like an easy fix, but companies also want ways to confirm the repair was made, guidance on what to do with the old part and answers to how it broke in the first place.

“Being able to manage that entire process and all of the work around the reverse logistics is another really interesting opportunity that we’re going to see more and more of as we’re getting data out of these things that are out in the field,” he said.

An Opportunity To Impress

From a financial point of view, customer-service business models based on the IoT can extend revenue beyond the initial point of sale and traditional maintenance and repair contracts, according to a February report from PwC.

The report, which assessed what the Internet of Things means for U.S. manufacturing, stated that companies may also build customer loyalty by spotting an imminent problem and performing “predictive maintenance” before a product breaks down.

This kind of technology is now being offered as an added service for car buyers. In vehicles equipped with OnStar, for example, drivers involved in an emergency can hit a button to get help, but the system also tells people involved in the response “everything the car is telling us,” said Schuerman.

Pegasystems has worked closely with OnStar in developing its customer experience systems.

“So we know what the car knows about what its last speed was, if it stopped, was there potentially an accident?” Schuerman said. “And that allows us to drive the right process in the service scenario that’s popping up.”

Start Somewhere

Sensors deliver an incredible amount of information, but most of the data isn’t actually used. On an oil rig with 30,000 sensors, only 1 percent of the data is examined for optimization and prediction, which provides the greatest value, according to a June report from the McKinsey Global Institute. The rest of the data is used mostly to detect and control anomalies that have already occurred.

Because many enterprises are generating vast loads of data, it’s imperative that adopters of IoT analytics technology choose a desired business outcome, said Colin Parris, vice president of GE Software Research. That way, they can dive into the data pool and surface with targeted intel. For example, a company trying to reduce the number of routine inspections may pull data related to the number and types of issues identified.

“If you can tell a customer how you can save them money, it gets them interested,” Parris said. “We’re not saying digitize everything.”

Data For Design

Data collected by manufacturers can also benefit research and development teams working on the next iteration of a product.

They can use analytics to identify frequent problems worth fixing or trends that may lead to tailored versions of the same product—such as an airplane engine ideal for shorter flights, or a locomotive engine specially designed to pull heavy freight, Parris said.

“This thing is a virtual loop,” he said. “The more data customers give back to us, the more value we can deliver to them. That’s the beauty of it.”

To learn more about evolved CRM, visit Pega.com.

Nick Clunn is an award-winning journalist who writes about consumer and information technology. Prior to working the tech beat, he covered news for daily newspapers in New Jersey, including The Record.