The rise of predictive maintenance is upon us

 The rise of predictive maintenance is upon us

In McKinsey & Company’s June 2017 report, “Artificial Intelligence: The Next Digital Frontier,” the institute writes about how artificial intelligence will induce a shift from preventative to predictive maintenance “in the future.” This is absolutely true. But it’s also true that this capability already exists. We don’t have to wait for it to arrive. Rather, we need to find the ideal implementation right now.

The potential of AI has been eagerly discussed for decades. But it’s time to abandon the futurist thinking that often colors the discussion and acknowledge that advanced AI is now a reality.

We are on the brink of a revolution led by Industry 4.0 and data science — a revolution that will transform every realm of the manufacturing industry while providing its managers with more control than ever before. The question, therefore, isn’t about whether players are going to ride the new wave that’s sweeping across the industries — it’s about whether they’re going to lead or merely fall in line.

The capabilities of Industry 4.0

As Industry 4.0 continues to push the boundaries of data science, the ability of AI to detect and signal precisely when maintenance is necessary will advance in kind. Contrary to the McKinsey report, companies are already utilizing data-driven insights to move beyond preventative maintenance. Instead of committing to arbitrary service schedules and the unavoidable waste, redundancy, and disruption that accompanies them, maintenance is handled dynamically.

Often, performance variables are tracked in real time using data from connected machinery. And when those variables indicate that performance is declining, a technician is able to step in before the asset malfunctions and shuts down an assembly line. Maintenance happens only when it’s necessary and always before it’s too late.

This kind of capability has been coveted for years by maintenance, repair, and operations personnel in industries ranging from manufacturing to oil and gas to pharmaceuticals to retail. But the wait is over.

General Electric has already created a technology that allows companies to model “digital twins” of actual machines. The performance of the twin is tracked closely, offering monitors an in-depth look at an asset throughout its life cycle. If and when an issue arises, it’s visible in ways never before possible. Followers of the McKinsey study’s view on waiting for the future of AI may be surprised to learn that 650,000 of these “twins” are already in operation.

Elevator manufacturer Schindler is another company leveraging predictive maintenance. The company makes most of its revenue providing service. By installing sensors throughout an elevator, the company is able to justify the need for every service appointment while improving the safety of elevators overall.

The importance of predictive maintenance

The arrival of predictive maintenance may not represent an exciting new capability, but it does open possibilities in which industries and verticals can grow. Let’s take a look at a few burning issues in industries where waiting around to use predictive maintenance could pose great risks:

  • Transportation and utilities: New York Governor Andrew Cuomo declared the Metropolitan Transportation Authority to be in a state of emergency, citing that it relies heavily on emergency issue resolution as a result of insufficient preventative upkeep. Unfortunately, this state of asset disrepair is not confined to a season, geography, or industry. Infrastructure across the U.S. — including highways, bridges, and sewers — is in dire need of repair. And as infrastructure continues to age, the maintenance burden will surge. Knowing where, when, and why maintenance is necessary is crucial for controlling costs while avoiding disasters.
  • Automakers: Recalls are extremely costly for manufacturers and quite dangerous (and annoying) for consumers. Just this year, General Motors recalled 690,000 trucks in the U.S. and about 100,000 in Canada and other countries. But if automakers relied on data and machine learning throughout the design phase and after the roll-out, they could uncover issues that require attention sooner rather than later.
  • Oil and gas producers: The U.S. relies on 2.5 million miles of pipelines to transport oil and gas — and more than half of those miles were constructed in the 1950s and ’60s. As these resources get older, the size and frequency of failures will only increase. Being able to predict and prepare for these issues in advance is invaluable.
  • Airlines: The Federal Aviation Administration handles 43,864 flights on a daily basis. And given the scope of the aviation industry, the dynamics of maintenance are too vast and too complex for any mind (or team of minds) to comprehend. Switching to a predictive maintenance strategy offers a simpler, safer, and more economical approach to maintenance that better serves the interests of everyone.

Making predictive maintenance a reality

Predictive maintenance is technically possible, but what is more significant is that it’s technically accessible. Companies do not have to make massive investments into unproven technologies in order to become proactive about maintenance. Rather, they just need to begin capturing, storing, and utilizing data that is already there.

The first step is to connect machinery to sensors that can track data in real time. That data should then be fed into a network that has the capacity to store and process large volumes. Finally, intuitive automation should be employed to track patterns and identify anomalies. If and when attention is required, a technician is deployed to perform corrective action that delivers the maximum ROI.

Predictive maintenance is an impressive tool to have in your kit. Companies that minimize maintenance costs, optimize field resources, reduce warranty claims, and insulate themselves from risk are at a significant competitive advantage. And the best news? The future for predictive maintenance is now. So what are you waiting for?

Sundeep Sanghavi is the cofounder and chief executive officer of DataRPM, a Progress company.

Let’s block ads! (Why?)

Big Data – VentureBeat