5 ways predictive maintenance cuts down on costly downtime

Manufacturers are urged to consider the ways in which data can help them remain innovative and ahead of a challenging marketplace. This could include predictive maintenance.
Manufacturers are being encouraged to utilise data capture to improve their approach to maintenance.

The industrial sector is undergoing massive transformation like never before. Sundeep Sanghavi co-founder of DataRPM looks deeper than preventative maintenance into the realm of predictive maintenance.

According to a World Economic Forum study, if the first Industrial Revolution was driven by steam, the second by electricity, and the third by digitization, we are now in the midst of a fourth: one driven by data. In this “Industrial Internet of Things” (IIoT) era, an explosion of IoT-connected devices has created a new level of big data possibilities for enterprises of all types.

While many consumer-facing industries such as retail and healthcare make bold attempts to keep up with the changing times, manufacturing generally lags behind in capitalizing on data-driven strategies. Granted, because of the technical complexity of data analysis and the ongoing shortage of data scientists, gaining meaningful business insights from the IIoT is far from easy — but that doesn’t need to be the reason manufacturers fall behind.

In the search for how and where manufacturers can best reap the rewards of the IIoT, look no further than maintenance protocols.

Prevent or predict?

For much of the modern manufacturing world, the prevalent approach to IT maintenance can be best described as “preventive” — performing maintenance on a schedule, even if machines don’t actually require updating or repair. According to a 2015 study, 85% of companies utilise this strategy.

This is a non-optimal approach because scheduling maintenance leads to unnecessary labor (and, therefore, unnecessary costs). Additionally, a potential problem that goes undetected between scheduled maintenance can pose huge risks to the financial health of the company and the physical health of its workers.

Currently, manufacturers tend to devote their technology toward creating and enforcing maintenance schedules — but now’s the time to begin evolving toward a predictive methodology.

The biggest perks of predictive maintenance

Using technology to drive a predictive maintenance approach, manufacturers can stream data from sensors mounted on their machines to uncover key usage and performance patterns in real time. This allows them to identify potential malfunctions in the making and avoid the hefty costs of unexpected breakdowns.

As a whole, predictive maintenance offers manufacturers a tremendous opportunity to boost operational efficiency. The process can be easily automated through machine learning and cognitive data science, meaning there’s no need to hire an army of data experts or assign additional human capital to the cause.

Here are five specific benefits manufacturers will see from a predictive maintenance approach:

  1. Reduced wrench time: Since the turn of the millennium, American factories have spent trillions of dollars a year performing maintenance. A predictive approach would render a great portion of these costs unnecessary or completely avoidable. For example, manpower would no longer be wasted on routine machine diagnostics, and companies could avoid huge outlays purchasing and installing new machines to replace those that broke down between inspections.Human intervention will be required only for a necessary repair, and because the issue is caught before it results in a complete mechanical breakdown, the wrench time will be much shorter.
  2. Less downtime: “Downtime” is a word manufacturers despise. About 5% of the average factory’s production is lost to downtime every year, costing the global manufacturing community roughly $647bn annually. However, with a predictive, data-driven maintenance approach, companies are much less likely to be blindsided by equipment failure and face lengthy, costly periods of inactivity. As an example, robotics company FANUC partnered with Cisco to create a “zero downtime” program that ended up saving one auto manufacturer $40m.IoT sensors can detect if a machine is beginning to break down and automatically order a new part. Then, the company can schedule the necessary repair or installation during a low-volume or after-hours window, minimizing lost productivity.
  3. Improved safety: Manufacturers undoubtedly take pride in creating safe environments for their workers, but one study found that up to 30% of all manufacturing deaths occur during maintenance activity. A predictive approach to maintenance does wonders toward increasing workplace safety by preventing dangerous scenarios. Companies can set safety parameters and receive immediate notification when a machine shows preliminary signs of reaching that threshold.
  4. Better inventory management of parts: By predicting what will fail, one can manage what parts he needs to order or have in his maintenance inventory.
  5. Better field personnel management: By predicting what will fail, manufacturers can better plan what kind of expertise would be required to perform maintenance. Furthermore, they can better plan schedules, too (rather than when ad hoc failures occur).

The lowered costs, higher productivity, and improved product quality made possible by predictive maintenance all lead to the ultimate benefit of happier customers. And further, employing this strategy arms companies with cutting-edge technologies empowered by cognitive data science that produce high-level business insights — unlocking countless opportunities for improvement and innovation.

Preventive maintenance can only carry you so far; join the next revolution by embracing predictive maintenance.

About the author: Sundeep Sanghavi is a highly accomplished data junkie, innovator, and entrepreneur with more than 20 years of experience in using data as the currency to perform advanced analytics. Known for his “what if?” mindset, he co-founded DataRPM with the goal of providing a platform that delivers hyper-fast data products to organizations challenged by the volume, velocity, and variety of their big data and machine learning.