Bringing Predictive Maintenance Home To Your Plant

What options are available? We asked SKF’s Andy Hoy about his company’s new service-program offering.

By Jane Alexander, Editor

MT: While the potential benefits of predictive maintenance (PdM) programs for machinery have been widely recognized, some operations have yet to move in that direction. Why?

HOY: A predictive maintenance strategy that lets operations identify and repair machinery problems before they escalate to full failure modes certainly can help any plant reduce the costs of maintenance, prevent breakdowns and unplanned downtime, increase machinery availability, improve productivity and limit production losses. But it’s true that despite the advantages, successful rollout and implementation of a predictive maintenance program sometimes can be thwarted by economics, logistics and/or a variety of other factors. In the case of smaller-scale operations, for example, the investment required for start-up equipment, training and initial support—not to mention the time and costs associated with ongoing analysis and reporting—may just be too much to carry.

MT: What options do such operations typically have?

HOY: For one, a plant can create and run a predictive maintenance program on its own, but at what costs?  Dedicated specialists must be hired, trained and retained; condition-monitoring equipment, including data collectors and related infrastructure, must be onboard; software and network systems to manage and analyze machine condition data must be installed; and maintenance resources must continually be enlisted to act on the generated data. All these elements may prove too daunting, especially for limited-scale manufacturers with an eye on the bottom line. In fact, based on historical data, costs for predictive maintenance programs can easily exceed $100,000 for start-up alone when run entirely in-house.

Another conventional way to go is to outsource a predictive maintenance program, but this can be tricky, too. Machinery operations must be scheduled around a contractor’s visits; access must be provided to all critical plant machinery; escorts may be required; “call outs” will have to be handled definitively in a timely manner; and confusion may develop regarding “ownership” responsibility of the particular maintenance effort.

MT: Is there another way to go to circumvent these issues?

HOY: Yes. We looked long and hard at the manufacturing landscape and confirmed that small- to mid-sized operations have been at a distinct disadvantage when it comes to establishing and running effective and affordable predictive maintenance programs. So we took steps to leverage our expertise and make it easier by developing and offering the SKF Machine Health Reporting Program (MHRP).
This program is particularly appropriate for operations with the following needs/characteristics: established goals to reduce maintenance costs; production requirements that must be achieved; up to 500 critical and interdependent rotating production machines and equipment for which high repair or replacement costs can be expected.

MT:  How does the MHRP stand out from conventional predictive maintenance programs?

HOY: The MHRP is very much about positively managing risk and involving a plant’s existing labor force in partnership with SKF. It’s a vibration-based maintenance service program engaged through our authorized distribution network and designed to build on the strengths of each partner—i.e., SKF expertise in maintenance strategies and predictive maintenance and a distributor’s inherent knowledge about customer operations and on-site logistics. We contribute the enabling technologies and expertise to collect data about the health of machinery at a facility and deliver reliable analysis, reporting and remedial recommendations. Equipped with ample warning, operations can get ahead of the curve on problems and take proactive measures to prevent catastrophic machinery failure, which is what predictive maintenance programs are all about.

MT: How does your program handle enabling hardware requirements?

HOY: In a way, this program is structured similar to a cellphone subscription, where the operation signs up for the “service plan” and SKF provides state-of-the-art vibration data collectors as a part of the “contract” and instructs the operation’s front-line workers on their proper use. Since these data collectors represent big-ticket items, we’ve found that this hardware contribution helps break the financial ice.

The hand-held SKF Microlog portable data collectors/FFT analyzers can capture full-feature route and non-route dynamic (vibration) and static (process) measurements from many sources. And the technology can be applied to any rotating equipment, including motors, gearboxes, fans, compressors and conveyors, among others. Signals from connected sensors are digitally recorded and stored and then uploaded for post-processing purposes, including analysis and reporting. The result is that an operation learns what’s wrong with a machine, the extent of the problem and what to do about it.

MT: How is data relayed and handled, and at what costs?

HOY: Analysis and reporting of data typically would necessitate the purchase of expensive software, installing it on servers maintained by an IT support group and preserving data integrity. But the MHRP takes the ball and runs with it. All can now be performed remotely with this program by taking advantage of SKF “cloud-hosted” software infrastructure, supplied software and analysis/reporting protocols.

In a nutshell, machinery information and measurement data are uploaded to the “cloud” server via the Internet, where it is stored in SKF @ptitude Analyst Software available for viewing any time and anywhere from Internet access on the ground. Incoming data are reviewed continuously by SKF @aptitude Decision Support, which automatically compares the new data against known (and good) baseline measurements for possible deviations. The software flags problems and alerts a designated SKF engineer, who reviews the data and decides on the best course of remedial action(s). A report follows with recommendations.

MT: If an operation signs up for the MHRP, what’s the timeline to get up and running?

HOY: In the very first month, an operation supplies a list of critical machines for scrutiny. Then, SKF comes on site and instructs staff on predictive maintenance fundamentals, collects machine information and builds a measurement database. In the second, or “launch,” month, the Microlog data collectors are delivered, instruction is provided, communication software is installed, baseline data is collected and the first in a series of Machine Health Reports is published. Once up and running, the program subsequently collects data and delivers Machine Health Reports monthly with quarterly on-site analysis meetings and on-demand “spot” checks included.

MT: Any success stories so far?

HOY: Absolutely. For example, a roofing-shingle manufacturer initially tried to implement a predictive maintenance program on its own, but found the process to be a rocky road: The operation’s busy in-house plant maintenance team couldn’t keep up with the program, inexperience with vibration analysis made it difficult to convert data into actions and frustration mounted. With machine reliability a top priority, the site turned to SKF’s MHRP.

The experience and results were impressive. Previous failures of a mission-critical compressor had been costing more than $30,000 in repair parts alone—and unplanned production downtime was measured in days, not hours. With our MHRP in full swing, it was determined that a failing bearing was at fault. By proactively replacing that bearing, the plant saved tens of thousands of dollars in the direct cost of repair parts. All work was completed during a regularly scheduled production stoppage with no additional interruptions experienced or necessary.

Other machinery was probed, too, including a critical pump and mill transfer blower, whose Mean Time Between Failures were too early and too often. Within the first six months of the MHRP approach being implemented, the plant was able to get rid of more than 25 hours of unplanned downtime and gain tens of thousands of dollars that had previously added up in lost productivity and repairs.

These cases demonstrate that the intrinsic MHRP partnership ultimately can position plant operations solidly on the road to realizing how a comprehensive predictive maintenance program can make the big difference, especially when supported by experts equipped with the knowledge to deliver sustainable success and savings. MT

Andy Hoy is Director-North American Machine Health Reporting Program at SKF USA Inc. Email: andy.hoy@skf.com.