When a predictive maintenance (PdM) program doesn’t go right, it’s usually not for lack of trying. In last month's issue, Part 1 of this story uncovered common flaws and recovery approaches for programs that fail, stall, underperform, or are defunded. The causes of failure usually boiled down to weakness in one or more critical success factors.
The sheer number of identified critical success factors, all interdependent and equally important, may come as a surprise to those who had considered PdM a “simple” reliability improvement initiative. In reality, the failure of any one element has the potential to put the entire program at risk.
To help you reboot and get it right the next time, following are seven critical success factors as explained by industry professionals who have either “been there, done that” or seen it happen. Their personal experiences and recommendations for restarts are well worth noting.
1. Make reliability a shared vision
Make the commitment from the onset and let everyone know that PdM is part of the way your plant does business, suggests Greg Padesky, SkillPoint account manager at Advanced Technology Services (ATS). “Without this kind of commitment, many naysayers will help ensure even a well-considered PdM plan fails.” He believes a cultural shift is required from the production floor to the corner office for a PdM program to receive the focus and follow-through required to yield long-term reliability and throughput gains.
How a program restarts really touches on the cultural and behavioral aspects of the organization. Ron Bitely, global E/I reliability manager at Arizona Chemical, a Kraton company, recommends developing a strong vision and alignment to your “North Star” – a common goal shared by all members of the team or organization. “The current state did not happen overnight, and you need to continually follow up to get it back on track. It’s all about being an ambassador for reliability.”
2. It’s about people, not just the tools
Yes, there needs to be an investment in tools or systems to collect and trend the data, but technology is only part of the solution. Someone trained in the fundamentals of PdM is needs to facilitate the program.
Larry Hoing, senior manager of asset care at Wells Enterprises, says predictive technology alone will not advance your abilities. “I have seen programs stall or fail because they do not have the talent or knowledge to use the technology. The how, when, and where to utilize the technologies is all a part of being successful.”
Sometimes programs fail simply because the reliability technician wears too many hats. “Pulling a person from the day-to-day battle of emergent work and giving them responsibility over only predictive work will in the long run pay off and help solidify the value of the technology,” suggests Hoing.
Building a team starts at the top. “Management must have the foresight to assign specific personnel to support the program, and then support them as they support the program,” says Bernie Blair, thermographer at International Paper. He suggests developing and training a set of inspectors and developing very specific inspection processes, preferably using a PdM management system.
3. Think like an analyst and act on the data
Dedicated time reserved for planning, inspection, and data analysis is vital to PdM success. In a really world-class PdM program, the team brings more to the table than just picking out machine defects, remarks Tim Dunton, director of reliable manufacturing at Reliability Solutions. They think like analysts, provide troubleshooting support, and get to the real root cause of equipment failure. If 90% of an analyst’s time is just routine, day-to-day data collection, then the program can never realize its full value.
ATS’ Padesky believes that predictive technologies may help identify equipment issues before the problems become catastrophic, but too often these findings are ignored because the equipment "looks fine." It is challenging to stick with the routine data collection required of a good PdM program if the findings are disregarded and not corrected.
4. Educate and empower the leaders
Like all human endeavors, PdM programs succeed and fail for one reason only – the strength (or weakness) of leadership, says Burt Hurlock, CEO of Azima DLI. “Successful PdM programs are the work of strong leaders who (1) set clear objectives, (2) measure their progress toward those objectives, and (3) institutionalize practices and beliefs that make the program sustainable through generational and organizational change. The effort to start or restart successful PdM programs begins with identifying leaders with the managerial skills to build and institutionalize the program.”
Reliability Solutions’ Dunton believes in leadership by example. “In our leadership training series, I make the leaders do exactly what I want their people to do, which is to document the case histories and show the return. It’s very powerful. With a class of 15 people over 18-24 months, we’re averaging between $4-6 million per class in documented, hard number savings from applying what we taught.” Once these leaders see the importance of showing justification, they go back and start demanding it of their people.
5. Plan ahead for champion turnover
When PdM succeeds, it’s usually the work of a rising star and the results are visible. Those champions are often drawn to higher, more lucrative positions, says Azima DLI’s Hurlock. Without their leadership, managerial skills and discipline, their PdM programs will flounder unless they have introduced a culture of PdM, in which case individuals and teams following in their tracks will manage to keep the program running.
“In industry today, with the desire to cut headcount and costs, very often it’s the PdM group that doesn’t get the succession planning they need,” observes Reliability Solutions’ Dunton. “It can take several years to get a person to the point where they are really adding value.”
Also, outsourcing can prove to be an effective solution when normal personnel turnover rates create a situation where there is no one trained to run the system or analyze the data, says Joe Van Dyke, vice president of operations at Azima DLI.
6. Communicate the value clearly and persistently
The lack of frequent communication and celebration of PdM benefits and metrics – from the maintenance team on up to senior management – causes support to lapse, says John Bernet, vibration specialist at Fluke. Every potential failure caught and corrected before causing unplanned downtime needs to be reported and celebrated.
“Producing regular business-level metrics reports on program performance, including avoided costs and ROI, can accurately account for the program savings and benefits and establish the historical track record that justifies the PdM program under budgetary pressures,” explains Azima DLI’s Van Dyke.
Wells Enterprises’ Hoing recommends making the results tangible and visible. “Put the bearing you found with vibration analysis in the hand of the director of operations, and explain how you found its defect and what failure was avoided because of using this technology. Show the IR image of the motor contactor with one leg having a 100 degree F delta from the others. Show the wear debris analysis of the oil sample taken from a critical gearbox.”
If the senior leadership doesn’t see the value, the program goes down, remarks Reliability Solutions’ Dunton. “In the old days when I was setting up programs, we used to teach the analysts how to make presentations in PowerPoint. Each analyst was required to make quarterly presentations of case histories to the site or corporate leadership and show the value. The art of putting that type of presentation together has declined today, but it remains effective.”
7. Fight for your successful programs
When a program is successful, maintenance overhead goes down, failures largely cease, and everything looks good. If new management isn’t aware that all of that success depends on PdM, they may see the PdM expenses as unjustified, cautions Fluke’s Bernet.
In the first two or three years of a PdM program, numerous critical machinery issues are usually identified, but as the program matures, these critical issues are encountered less frequently because the root problems are identified at an earlier stage and rarely progress to criticality, explains Azima DLI’s Van Dyke. “When looking for ROI, these relatively mature PdM programs are often seen as a cost that is only identifying relatively minor asset issues," he says. "This ignores the fact that, in the absence of a PdM program, these minor issues would inevitably turn into big and costly issues.”
Business ownership changes, such as mergers and acquisitions, similarly put PdM program support at risk. “The buyer needs to be educated upfront that the performance they’re buying is the result of a disciplined machine condition monitoring program,” says Azima DLI’s Hurlock. “Otherwise, they may terminate the program in the belief they’re saving money, only to begin incurring large and unplanned maintenance and capital expenses soon thereafter, usually within two years.”
Frequent program justification is needed, vocally and with documented proof points, to ward off anyone with the potential to weaken or defund programs that have demonstrated continuing success.
Even if all the right steps are taken, be prepared for pushback, cautions Padesky. “Many plants have already tried to utilize PdM, and most have failed due to a lack of organizational ownership, lack of follow-through, and lack of vision. Communicate your plan, document your wins, and stick with it.”