Companies implementing predictive maintenance (PdM) face a variety of challenges. Though the technical aspects will garner the most attention, human nature plays a strong role in the program’s success. Recent discussions with practitioners and providers shed some light on lessons learned.
1. Usability: Today the challenge lies not necessarily with explaining the value of a PdM approach, but in making it easy for organizations to use, says Paul Lachance, president and CTO at Smartware Group. He suggests asking yourself:
- Where is the data? You will need it to collaborate with your teammates.
- What are the thresholds to initiate a PM? Confer with the experience of your team, vendors, historical CMMS data, and others.
- What actions should you take? This will require assistance from CMMS, teams, and vendors.
2. Executive support: Effective collaboration comes from strong leadership. “Until reliability became strategic, PdM rarely received the managerial sponsorship required to deliver its full potential. That is changing, albeit slowly,” remarks Burt Hurlock, CEO at Azima DLI.
3. Talent: PdM programs are notoriously inconsistent, so companies are looking for ways to retain talent or rely on third parties to maintain PdM disciplines, says Hurlock. In that sense, PdM can often be a good barometer of company culture. Those who make long-term commitments to sustainability, however modest, tend to hold on to their gains while the ones who don’t find themselves continually resetting.
4. Trust: Whether it’s in the cloud or local, companies are still learning to trust PdM data. “Management will have greater confidence in data that comes directly from the tool rather than an ear-pencil-transcription process. They’ll know who took the data, when it was taken, and that it is traceable and directly aggregated,” suggests John Neeley, product planner and program manager for Fluke Connect Assets.
5. Resistance: The older generation is not quite as adaptable to the new tools and technologies. “It goes back to the psychology of change; you’ve got to show people what’s in it for them and involve them in the conversation,” remarks Tim Dunton, director and instructor/developer at Reliability Solutions.
6. Vision: Corporations are beginning to recognize reliability as a significant competitive advantage, so there is less willingness to share information outside of their own organization, Dunton observes. Some may even have rules to prevent it from happening. “I personally think that’s backwards thinking, because if you want to improve, you have to have targets, and you have to know what ‘good’ looks like,” he explains.
7. Interaction: “Software and technology providers have certainly put a lot of work into giving us the ability to share and exchange information, but don’t forget the psychology of sharing. When everything’s up on a network somewhere, then the people don’t actually get together anymore. There’s huge value to having your PdM techs have conversations about the problems and get different perspectives,” adds Dunton.
8. Context: When virtual experts are used, make sure they truly understand the dynamics of the asset that’s running, because otherwise their advice may be misleading. “If a conveyor designed for one amount is now running double or triple that, it could lead to different signatures being picked up on the technology. The expert might call a defect or miss a defect when a machine is running under a different operating context or in a different environment,” says Shon Isenhour, partner at Eruditio.
9. Methods: “It is important to use the same methods and nomenclature, including equipment names and associated asset numbers, when setting up routes across all of a company’s facilities,” suggests Adrian Messer, manager of U.S. operations at UE Systems. “It seems simple, but this is one thing that most companies fail to do when implementing a PdM technology like ultrasound across multiple facilities.”
10. Standards: Several standards for communication are vying for wide adoption. Once major equipment manufacturers adopt truly open standards for their sensor communications, we'll see significant expansion in the use of machine sensor data usage, remarks Mary Bunzel, portfolio manager for IoT, EAM & Analytics at IBM.
11. Security: “The largest obstacle we have to overcome is the acceptance of cellular and cloud technology,” says Paul Berberian, condition monitoring specialist at GTI Predictive Technology. “Some IT departments won’t allow cloud access to the plant network. One workaround is to offer alternatives, such as storing the data on an iPad, in a Dropbox, or sending it via email and viewing the data on a local viewer.”