State of the Industry: PREDICTIVE MAINTENANCE€¦ · State of the Industry: PREDICTIVE MAINTENANCE...

www.PlantServices.com-1-

State of the Industry:PREDICTIVE MAINTENANCE

MARKET RESEARCH REPORT


CONTENTS

PdM Balance 3

Asset monitoring on the go, Part 1 9

Asset monitoring on the go, Part 2 11

MARKET RESEARCH REPORT

By Thomas Wilk, Editor-in-Chief

Our 2017 predictive maintenance survey reveals

how plant teams are tying fresh approaches into their PdM programs.

PLANT SERVICES: MARKET RESEARCH REPORT


Read on for the survey highlights,

and then download the full set of 2017

PdM survey data at plnt.sv/1704-PDM.

DECISION-MAKING AUTHORITY AND

PRESCRIPTIVE ANALYTICS

Data from the first two questions

of the survey – what is your job

function (see Figure 1) and how

many plants does your organization

manage (see Figure 2) – demonstrate

further progression of the trends

identified last year.

A plurality of 2017 respondents

(22.5%) still identify as maintenance

managers, with reliability engineers

as the next largest category (15.7%).

Technicians were more strongly repre-

sented this year in both maintenance

and reliability, and several non-MRO

job categories continue to hold steady

over time (controls and applications

engineers, sales and marketing, and

executive-level).

The decline of traditional “plant”

job titles continues, as these posi-

tions seem to be evolving into one of

the more-specialist categories listed

under both maintenance and reliabil-

ity. However, it is worth noting that

you also told us that decision-making

power over monitoring tools and con-

figurations still resides primarily in

the hands of the plant manager (37.5%)

Last year you told us that, despite planned and actual

investments in predictive maintenance, you were not all that

happy with the results of these.

What a difference a year can make – more PdM technolo-

gies than ever are being deployed, often in targeted and

complementary fashion, and your levels of satisfaction are

on the way up. Even the challenges to success have shifted

away from how to define expected benefits and more to

whether you have enough people (and the right people) to

execute effectively.

WHAT IS YOUR PRIMARY JOB FUNCTION? 2014 2017

Plant manager

Plant engineer

Maintenance manager

Maintenance engineer

Maintenance technician

Reliability engineer

Reliability technician

Controls engineer

Sales/marketing

Applications engineer

Corporate executive

figure 1

WHICH TYPES OF MAINTENANCE APPROACHES ARE CURRENTLY EMPLOYED BY YOUR PLANT?

Reactive (run-to-fail)

Preventive (periodic and/or route-based)

Predictive (diagnostics ahead of failure)

Prescriptive (diagnostics and guidance for repair) figure 3

12.8%6.7%

15.2%4.5%

24.2%22.5%

7.1%12.4%

7.1%7.9%

9.0%15.7%

7.1%12.4%

5.2%5.6%

1.4%1.7%

3.8%3.9%

6.6%6.7%

HOW MANY TOTAL PLANTS DOES YOUR ORGANIZATION MANAGE? 2014 2017

figure 2

1 plant

2-5 plants

6-15 plants

More than 15

35.9%

21.9%

29.3%

31.5%

13.7

16.3%

21.2%

30.3%

40.5%29.8%

15.3% 14.5%



and the maintenance manager (57.9%),

despite the fact that significantly fewer

of these job roles seem to exist now

than in 2014. Some decision-making

authority does rest with the reliability

engineer (25.0%) and maintenance

engineer (23.0%), perhaps reflecting

that those job titles are on the rise.

It’s useful to consider these data in

the context of the trend in total num-

ber of plants organizations are manag-

ing. Figure 2 spotlights the continuing

trend toward plant consolidation, with

the number of single-plant operations

having dropped by almost 40% since

late 2014, and the number of opera-

tions with 15 or more plants having

increased by 43%. When contrasted

with job title data, it suggests that

plant managers are often being asked

to take responsibility for more than

one plant at a time – a finding that was

reinforced by our 2017 Workforce sur-

vey, in which 47% of respondents told

us that their current responsibilities

include additional plants beyond their

primary location (see our February

2017 cover story).

For this year’s survey, we also want-

ed to ask about how and your teams

are approaching the Next Big Thing:

prescriptive analytics. If predictive

maintenance is concerned with what

is likely to happen to an asset based

on a given set of condition monitor-

ing data, prescriptive analytics is the

logical next step, using those data sets

as well as other less-traditional data

sets (such as MES, ERP, and/or GIS

data) to identify a set of actions or

decision options that will address the

earlier prediction.

In essence, prescriptive analyt-

ics extracts actionable insight from

piles of data, answering the question,

“What is the best course of action for

a given situation?” Also, as machine

learning and predictive modeling

offerings get more sophisticated (and

affordable), prescriptive analytics can

continually take in new data to re-

predict and re-prescribe, thus learn-

ing with plant teams by improving

prediction accuracy and prescribing

better options over time.

This year’s PdM asked two brief

questions to identify where on the

prescriptive maintenance spectrum

you and your industry peers would

place yourselves (see Figures 3 and 4).

The good news is that almost half of

respondents (45.1%) indicated they

were engaging in some form of proac-

tive maintenance approach, whether

predictive or prescriptive. Of those

who indicated they were exploring

prescriptive options, 37.4% said they

were engaged now with prescriptive

maintenance, and an additional 10.7%

said there was money in this year’s

budget for prescriptive. (Just under

30% of respondents indicated that they

have no plans in this direction.)

TYING PdM PROGRAMS TOGETHER

In one of the more discouraging find-

ings from last year’s survey, readers

indicated that their overall levels of

dissatisfaction with their PdM pro-

grams had increased, despite planned

modest growth in levels of PdM in-

vestment. (Three areas of investment

in particular stood out as gaining

since 2014: control systems, EAM/

PdM PROGRAM: PERFORMANCE COMPARISON, 2014-2017

2014 2016 2017

Not effective 15.5% 15.6% 8.4%

Needs some improvement 40.3% 49.4% 45.8%

Satisfactory 24.8% 18.2% 21.4%

Effective 15.5% 14.3% 17.6%

Very effective 3.9% 2.6% 6.9%

figure 5

figure 4

10.7%23.7%

29.8% 37.4%Using now

In 2017 plans/budget

Within 3 years

No plans

TO WHAT DEGREE DOES YOUR TEAM OR ORGANIZATION CURRENTLY ENGAGE IN PRESCRIPTIVE MAINTENANCE?



CMMS systems, and predictive analytics software.)

We are happy to report that things have changed: The overall

level of satisfaction with your PdM programs has gone up by

34% since our last survey, and the share of respondents who

thought that their PdM programs were “very effective” nearly

tripled. Figures 5, 6, and 7 include trending data from our past

three surveys, from 2014-2017, to better illustrate this rise in

satisfaction and showcase the organic growth in application of

PdM tools and technologies among survey respondents.

When asked which PdM technologies are being deployed,

respondents indicated that the same three options that led the

2016 survey (infrared, oil analysis, and vibration) also lead

this year’s survey, with each being selected by more than 70%

of respondents and with infrared again the most-deployed

PdM technology (at 74.8%). Two other data trends stand out:

(1) the general increase in PdM technology use, and (2) the

decline in reported use of predictive modeling software.

Ultrasound, electric motor testing, acoustic, and corro-

sion technologies registered increases in use, suggesting that

plant teams are more comfortable than ever with balanc-

ing a wide suite of PdM tools and techniques in their asset

management programs. This includes internet-enabled tools:

Figure 8 identifies the asset types on which plant teams are

using IIoT equipment – primarily automation and control

WHICH PdM TECHNOLOGIES HAVE YOU DEPLOYED?

Using now In this year’s budget Within 3 years No plans

2014 2016 2017 2014 2016 2017 2014 2016 2017 2014 2016 2017

Vibration 60.0% 72.1 70.5% 5.8% 7.0% 9.8% 12.9% 9.3% 7.6% 21.3% 11.6% 12.1%

Ultrasound 45.5% 52.3% 59.1% 5.2% 9.3% 12.1% 16.9% 17.4% 7.6% 32.5% 20.9% 21.2%

Acoustic 24.7% 27.1% 34.4% 6.5% 3.5% 5.3% 14.3% 22.4% 16.0% 54.5% 47.1% 44.3%

Corrosion 33.8% 35.3% 45.0% 7.8% 2.4% 3.1% 14.9% 22.4% 10.7% 43.5% 40.0% 41.2%

Infrared 65.8% 66.3% 74.8% 3.9% 7.0% 3.1% 15.5% 10.5% 7.6% 14.8% 16.3% 14.5%

Oil analysis 62.3% 75.6% 73.5% 4.5% 3.5% 5.3% 15.6% 4.7% 6.1% 17.5% 16.3% 15.2%

Predictive modeling software

17.5% 25.9% 20.6% 6.5% 7.1% 10.7% 25.3% 23.5% 19.8% 50.6% 43.5% 48.9%

Electric motor testing

50.0% 45.3% 47.0% 5.8% 9.3% 9.8% 14.9% 23.3% 21.2% 29.2% 22.1% 22.0%

PLEASE RATE THE OBSTACLES LIMITING THE SUCCESS OF YOUR PdM INITIATIVES

Not a factor Low Medium High

2014 2016 2017 2014 2016 2017 2014 2016 2017 2014 2016 2017

Budget constraints 5.8% 5.8% 6.8% 14.3% 16.3% 18.2% 37.7% 44.2% 40.9% 42.2% 33.7% 34.1%

Undefined financial benefits

9.1% 7.0% 10.6% 18.8% 24.4% 27.3% 46.1% 32.6% 39.4% 26.0% 36.0% 22.7%

Undefined operational benefits

16.9% 14.0% 9.1% 26.6% 22.1% 33.3% 39.6% 37.2% 45.5% 16.9% 26.7% 12.1%

Limited engineering resources

16.2% 11.6% 10.6% 22.7% 24.4% 22.0% 42.9% 39.5% 39.4% 18.2% 24.4% 28.0%

Poor program execution

24.0% 20.9% 16.7% 32.5% 34.9% 25.8% 30.5% 31.4% 40.9% 13.0% 12.8% 16.7%

figure 6

figure 7



assets (67.6%), electrical systems (31.8%), and rotational/me-

chanical assets (25.2%). Also, when asked the more-general

question of whether respondents were deploying internet-

enabled technologies at part of their PdM program, numbers

did not change significantly over the past 12 months, with

about 20% of current respondents indicating they are cur-

rently engaged, and 55.6% of respondents saying they had no

plans to deploy these technologies.

It came as a surprise that predictive modeling technolo-

gies declined in reported use by 2017 survey respondents;

after increasing in use by almost 50% from 2014-2016, the

reported use of predictive modeling declined by about 20%

over the past 12 months. This could possibly be the result

of the success of other PdM approaches, where the need for

predictive modeling might be reduced as teams experience

greater success with PdM technologies that deliver condi-

tion monitoring data in real time.

This also may ref lect the trends observed in Figure

7, which charts the stated obstacles to PdM success. Al-

though budget constraints continue to lead all obstacles

reported on the survey, several new trends are emerging.

Specifically, respondents are more confident in their abil-

ity to articulate PdM financial and operational benefits

(62.1% and 57.6%, respectively) than in previous years,

whereas the challenges of “limited engineering resources”

(67.4%) and “poor program execution” (57.6%) are start-

ing to take their place as top-of-mind concerns. In other

words, the PdM roadmaps are coming into stronger fo-

WHO USES THE INFORMATION PROVIDED BY YOUR PdM SYSTEMS AND WITH WHAT FREQUENCY?

Never Weekly Monthly Quarterly

2014 2016 2017 2014 2016 2017 2014 2016 2017 2014 2016 2017

In-house maintenance

8.0% 3.9% 6.2% 59.4% 55.8% 48.7% 20.3% 27.3% 33.6% 12.3 13.0% 11.5%

In-house operations

29.0% 24.7% 32.7% 39.1% 36.4% 34.5% 24.6% 23.4% 16.8% 7.2% 15.6% 15.9%

In-house reliability engineers

31.9% 16.9% 21.2% 40.6% 44.2% 44.2% 16.7% 24.7% 27.4% 10.9% 14.3% 7.1%

Totally outsourced 65.2% 59.7% 76.1% 8.0% 6.5% 6.2% 14.5% 16.9% 10.6% 12.3% 16.9% 7.1%

Third-party remote monitoring

73.9% 74.0% 76.1% 10.9% 3.9% 11.5% 4.3% 7.8% 8.8% 10.9% 14.3% 3.5%

OEM supplier 71.7% 77.9% 69.0% 8.0% 2.6% 6.2% 7.2% 6.5% 12.4% 13.0% 13.0% 12.4%

WHICH TYPES OF ASSETS ARE YOU USING INTERNET-ENABLED / IIoT TECHNOLOGIES TO MANAGE? 2016 2017

figure 9

Automation assets (field devices, control valves)

Control system assets (DCS, I/O, controllers, networks)

Production assets (rotating equipment, mechanical pumps)

Distribution pipelines

Electrical systems

Fleet vehicles

HVAC/R system

Safety systems / devices

Manufacturing productivity

figure 8

19.5%27.2%

20.8%40.0%

27.3%25.2%

6.5%7.9%

24.7%31.8%

9.1%6.0%

13.0%18.5%

15.9%

15.2%



cus, but will there be enough skilled staff to deliver on the

predicted promise?

Finally, the types of assets that readers tell us they are manag-

ing via PdM did not change significantly from 2014 to 2017; in

general, respondents consider their production assets the top

priority, followed by electrical, automation, and control systems.

WHO GETS TO SEE YOUR PdM DATA?

One of the more interesting trends uncovered by the PdM

survey over the past several years is the general unwillingness

of respondents to share their data with experts outside of their

organization. The 2017 survey is no exception – you’ve told

us yet again that if you can avoid sharing data with OEMs or

other third parties, then you will.

When it comes specifically to the ability to use remote

monitoring technologies to share data back with OEMs,

the share of respondents who said they were doing this now

dropped from 31.9% in 2016 to 18.6% this year; and the num-

ber of respondents with no plans to do so jumped from 38.5%

to 60.2%. These data aligned with a separate question which

asked whether respondents were using embedded remote

monitoring devices as part of their PdM program. Just shy

of 40% of respondents said yes, which suggests that users are

leveraging embedded devices to monitor the plants they are

being asked to manage remotely, and that they are not always

sharing these data with OEMs.

Figure 9 provides additional nuance to this issue of PdM

data-sharing, with respondents able to drill down into the types

of teams they share their data with as well as the frequency of

sharing. In general, in-house maintenance and reliability teams

have quite frequent access to the data, with most respondents

sharing it weekly (44%–48%) and/or quarterly (27%–33%).

The greatest reluctance on data-sharing is saved for third

parties: a whopping 76.1% of respondents say they have no

plans to share PdM data with third-party remote monitoring

service providers, and 69.0% say they have no plans to share

PdM data with OEMs. The sweet spot, such as it is, seems to be

monthly data sharing, with 12.4% of respondents sharing with

OEMs and 19.4% sharing with other contracted third-parties

(similar to the rate of monthly data-sharing with opera-

tions teams). It’s unclear whether the resourcing or program

execution challenges identified in this year’s PdM survey will

eventually soften the overall resistance to sharing PdM data

outside the organization.



Condition Monitoring on the Go, Part 1New portable solutions let you monitor machines from anywhere – even your couch

Portable tools for condition data collection and moni-

toring provide an efficient and economical alternative to

earlier approaches. Portability gives equipment-facing

personnel in operations and maintenance an opportunity to

capture and act on changing conditions as needed – at any

time, from anywhere.

ON-SITE CONDITION REPORTING

Portability is ideal for addressing unexpected or intermit-

tent conditions as well as routine concerns on the plant floor.

The FactoryTalk TeamONE smartphone app from Rockwell

Automation focuses on increasing worker productivity by

providing near-instantaneous incident and device data and

allowing for collaboration among plant-floor, engineering

and IT workers. Incidents are posted in the iOS or Android

app and shared with others for analysis; the identified solu-

tion is then posted for execution.

“We are able to drive a reduction in mean-time-to-repair

through team collaboration, live device diagnostics, and

interactive machine alarms,” says Kyle Reissner, mobility

platform leader at Rockwell Automation.

WALK-AROUND CONDITION MONITORING

Solutions designed for walk-around condition monitoring

let analysts spend less time collecting data and more time

analyzing it. GE’s Bently Nevada SCOUT200 Series is a rug-

ged system that features a lightweight hip-mounted wireless

vibration data collector (the intrinsically safe SCOUT220-

IS model or COMMTEST220 non-IS-rated model) that

streams data via Bluetooth to a durable Android smart-

phone. The system has full connectivity with GE’s System 1

software for diagnostics.

“Portables provide a solid foundation and are a valuable

component in any condition monitoring program,” says

Chris McMillen, GE distributed hardware product manager.

The small, military-grade, IP67-rated Auguscope from

Augury records vibration and ultrasonic sensor data for

mechanical diagnostics, leak detection, pump cavitation,

and steam-trap issues. The technician attaches a mag-

netic sensor to the machine’s body and initiates recording

from his or her smartphone. That data is sent to Augury’s

servers for analysis.

If you can use Facebook or Twitter, then you can use the

Auguscope, suggests Saar Yoskovitz, CEO and co-founder of

Augury. “The Auguscope app combines an extremely intui-

ROCKWELL AUTOMATION

http://plnt.sv/1702TT-01



tive user interface with cloud-based algorithms to deliver

real-time machine diagnostics and treatment recommenda-

tions to users’ smartphones,” he says.

The TRIO line of vibration data collectors and analyz-

ers from Azima DLI can be worn on a belt, carried, or

slung over the shoulder by technicians or operators.

TRIO features four simultaneous channels of vibration

data and a dedicated tachometer, plus wireless Bluetooth

acquisition and voice recognition technology that allows

routine or complex troubleshooting data to be collected

from a safe distance.

Putting the data in the cloud gets more people involved

in diagnosis and repair decisions. “Data is data; it is what

you do with it that really matters,” explains Michael

DeMaria, director of product management and training

at Azima DLI.

IMI Sensors manufactures a suite of accelerometers to

optimize walk-around vibration monitoring routes. The

sensors include 5000g+ shock protection to guard against

damage from accidental overloads during sensor mounting

and removal. Integral magnetic bases and/or cables mini-

mize sensor installation time at each individual route stop.

Coiled cables and breakaway connectors enhance conve-

nience and safety for the technician.

“IMI Sensors provides tools that facilitate expeditious

measurements while ensuring the safety of the technician,”

says Meredith Christman, product manager at IMI Sensors.

The OM-DAQXL Series data logger from Omega Engi-

neering is a handheld multichannel touchscreen data logger

available with eight or 16 universal inputs. It is suited for au-

tomation environments in which high-speed measurements

are needed – for example, in chemical, water treatment, or

HVAC process monitoring facilities.

In the plastics industry, for example, the OM-DAQXL can

be used to measure temperature sensors (thermocouples or

RTDs) and pressure transducers used in various types of

plastics processing machinery, including injection-molding

machines and extruders, says Anthony Corvini, DAS and

automation product manager at Omega Engineering.

MOVABLE CONDITION SENSORS

With Fluke Condition Monitoring, rugged wireless sensors

for voltage, current, temperature, and power can be easily

connected and left in place or disconnected and moved

where needed. A gateway receives the sensor signals from

up to 30 feet away and sends the measurements and alarms

to cloud-based software that is viewable from a computer or

mobile device.

“The Fluke Condition Monitoring system allows main-

tenance managers to see more of what is going on in their

plant without needing more people,” says Weishung Liu,

Fluke product planner. “They can even monitor from their

couch.”

Email Contributing Editor Sheila Kennedy, CMRP, managing direc-

tor of Additive Communications, at [email protected].

www.rockwellautomation.comwww.gemeasurement.comwww.augury.comwww.azimadli.comwww.imi-sensors.comwww.omega.comwww.fluke.com

REFERENCE WEBSITES:



Asset Monitoring on the Go, Part 2Mobile, multifunctional tools aim to be game-changers for asset health analysis

Asset condition monitoring devices are achieving new

levels of usability, making data collection, analysis, and

corrective actions more efficient and timely. Smartphone-

enabled solutions, multipurpose devices, and an innovative

handheld oil analysis tool allow anomalies to be spotted

earlier and addressed more effectively with collaborative,

intelligent maintenance decisions.

SMARTPHONE-SUPPORTED IR TOOLS

Low-resolution, smartphone-based thermal imagers can

be a powerful new tool for helping monitor assets. Putting

these NDT tools in the hands of operations personnel allows

anomalies to be identified and reported earlier and lets man-

agers plan and schedule minor repairs before further dam-

age occurs, says Rob Miller, reliability director at Stockton

Infrared Thermographic Services.

Attaching a miniature IR camera to an operator’s smart-

phone or tablet allows visual temperature information to

be collected and distributed to maintenance personnel im-

mediately via text or email, Miller explains. “The relatively

low cost of implementing this technology makes it easy to

integrate into existing daily routines,” he says.

To facilitate the collection and trend analysis of IR in-

spection data, new E Sentry Connect asset tags from IRISS

use near-field communication (NFC) Smart Card technol-

ogy that allows contactless communication with Android

smartphones and tablets.

E Sentry Connect tags with embedded maintenance

instructions are deployed on the assets. New readings and

detailed notes entered in an app on the smartphone or tablet

are then recorded onto the tag by “bumping” the tag, ex-

plains Rudy Wodrich, VP of engineering services at IRISS.

The recorded data is uploaded to the cloud the next time

the device connects to the internet, and from there it can be

transferred to an asset management (EAM/CMMS) system.

MULTIFUNCTION VIBRATION DEVICES

User-friendly portable vibration meters record, analyze,

and display vibration signals with a click of a button. The

one-click TPI 9071 smart vibration meter from Test Prod-

ucts International displays color-coded alarm levels for

ISO values and bearing damage units (BDU) readings. It

detects imbalance, misalignment, and overall looseness as

well as more-complex issues.

PRUFTECHNIK

http://plnt.sv/1708TT-01



“You can buy a vibration data collector with spectrum ca-

pabilities for less than a smartphone and much easier to use

than the smartphone,” says Jim Weidner, vibration special-

ist at TPI. “When you do more-frequent checks, you catch

impending problems such as water in the oil, overheating,

loose and worn belts, and low lubrication.”

A multifunction portable tool for performing route-based

vibration analysis of rotating machinery is the OneProd

Falcon Smart Data Collector from VibrAlign. Designed for

efficiency, it has built-in features including a digital camera, a

strobe light, a spot pyrometer, a bar-code scanner, voice anno-

tation, two-channel balancing, a wireless triaxial sensor, and

patent-pending Accurex automatic diagnosis software.

“Accurex uses proprietary technologies to diagnose

machine problems such as imbalance, misalignment, fric-

tion, gear defects, and cavitation within seconds, next to

the machine,” says Brian Shanovich, ACOEM product line

manager at VibrAlign. It also displays confidence and sever-

ity levels and offers suggested actions.

The Vibxpert II from Pruftechnik is a handheld tool for

efficient route-based data collection, vibration analysis,

and field balancing. It can also be used to perform ac-

ceptance measurements of newly produced or installed

rotating equipment. “One- or two-plane balancing and

advanced diagnostic tools such as FFT coast-down, bump

tests, or modal analysis can be used in the off-route mea-

surement and analysis mode,” says Florian Buder, CEO at

Pruftechnik North America.

Pruftechnik’s associated Omnitrend software allows users

to manage their assets based on machine-specific vibration

data. Optional automatic measurement location identifica-

tion with Pruftechnik’s Vibcode transducer is designed to

maximize data quality during route measurements.

PORTABLE INFRARED OIL ANALYSIS

Reliability engineers can conduct oil analysis on their

maintenance route via the FluidScan handheld infrared

oil analyzer from Spectro Scientific. FluidScan provides

quantitative measurement of dissolved and free water us-

ing infrared spectroscopy, explains Randi Price, senior ap-

plications chemist at Spectro Scientific. “This allows rapid,

on-site determination of water contamination for critical

pieces of machinery, such as turbines, without having to

wait for results to come back from a lab,” Price says.

FluidScan uses a built-in library of more than 700 oils

and reliable calibrations to accurately report key parameters

including oxidation, water, TAN, and TBN. It helps provide

an overall picture of oil chemistry and contamination.

Email Contributing Editor Sheila Kennedy, CMRP, managing direc-

tor of Additive Communications, at [email protected].

www.stocktoninfrared.comwww.iriss.comwww.testproductsintl.comwww.vibralign.comwww.pruftechnik.comwww.spectrosci.com

REFERENCE WEBSITES:


State of the Industry: PREDICTIVE MAINTENANCE€¦ · State of the Industry: PREDICTIVE MAINTENANCE...

Documents

Transcript of State of the Industry: PREDICTIVE MAINTENANCE€¦ · State of the Industry: PREDICTIVE MAINTENANCE...