- The most common goal of predictive maintenance is improved uptime, with 51% of respondents in a PwC survey last year saying this was their organization's reason for adopting the technology.
- Other reasons for implementation include cost reduction, improved customer satisfaction and a longer lifetime for assets. While only 1% of respondents said energy reduction was a reason for using predictive maintenance, 36% said they experienced energy savings as a result of using it, according to PwC.
- One of the ways in which cost reductions can be achieved is through lower inventory levels that result from the use of predictive maintenance. "Generally speaking, most companies carry way, way too many spare parts in their inventory," Pete Guarraia, the global head of supply chain for Bain and Company, told Supply Chain Dive.
Predictive maintenance can work in concert with technology like location tracking to change the relationship that companies have with their parts suppliers, Robert Schmid, a managing director and chief futurist with Deloitte Consulting, told Supply Chain Dive in an interview.
Companies are increasingly connected to their suppliers and are getting to the point where it's possible to automate ordering from them based on inventory levels, Schmid said.
Generally, beginning a predictive maintenance practice should include a reimagining of the spare-parts inventory and overall maintenance plan, Guarraia said.
"The best companies are able to drive meaningful cuts," he said. "And when I say meaningful, I'm talking about 25% to 35% reduction in spare parts inventory, and some even as much as 50% reductions in the spare parts inventory."
Predictive maintenance allows a company to focus its spare parts inventory on the most critical parts, those with long lead times and those subject to unexpected outages. The rest can be handled by the vendor, Guarraia said.
Getting these results, though, is not easy. "Conversations with many customers reveal that implementing predictive maintenance solutions has been more difficult than anticipated, and it has proven more challenging to extract valuable insights from the data," reads a Bain and Company report from last year.
Often, predictive maintenance will not be the first step in making a more connected factory, Schmid said.
"Often this goes into a phase two or phase three. Often we find we need to first spend time really centralizing the machines or looking at the processes and really go through that before we go to predictive maintenance," he said. "So while it's one of the most talked about [forms of IoT], it's not one of the most implemented."
Many who implemented predictive maintenance expected use of the technology to lead to an immediate 10% to 15% return, Guarraia said, but this isn't the case. Companies have to figure out what they want to measure and which pieces of equipment are mission-critical and would benefit from this kind of investment. Figuring this out requires finding the bottlenecks in a factory, knowing what pieces of equipment are creating those bottlenecks and what kind of failure those pieces of equipment are typically experiencing.
Once these questions are answered, then a path toward predictive maintenance can begin to be laid. But even then, optimizing the relationship between inventory and maintenance is no easy task. Researchers have worked to build models that help to make the connection between the two. If it's achieved, then the company could benefit from cost-savings related to lower inventory and inventory storage.
This story was first published in our weekly newsletter, Supply Chain Dive: Operations. Sign up here.