Power of Predictive Maintenance
Power of Predictive Maintenance
What’s driving
of predictive
automotive
traceability today?
maintenance
Four use cases
An expert guidedemonstrate how real-
to the underpinnings
time insightschanging
of a rapidly into legacyindustry
assets minimize
downtime and lower costs
Introduction
Unplanned machine downtime is a huge headache for the business sector. In 2016, the Aberdeen Group found
that these events cost companies an average of $260,000 per hour – a 60% increase over its 2014 survey. While
Aberdeen hasn’t published any further updates, it seems likely that ongoing problems – including pain points
like aging infrastructure, skilled worker shortage, and recent societal disruptions – are continuing to drive up
operational costs by reducing equipment availability.
In manufacturing, plant operators are tackling unplanned machine downtime by migrating from preventive
maintenance to predictive maintenance. Unlike preventive maintenance, which is performed according to
a schedule that reflects historical events, predictive maintenance monitors asset condition in real time and
prompts interventions before failures disrupt production. Advances in sensors, analytics, and communication
technologies are making predictive maintenance increasingly practical and affordable for small, medium, and
large manufacturing companies.
This white paper discusses the ways in which a solid predictive maintenance strategy backed by state-of-the-
art technologies can help manufacturers stay ahead of equipment problems and minimize downtime. Four use
cases are provided to help guide decision makers and define work efforts and budgets.
2
Contents
Summary...........................................................................................10
The power of predictive maintenance
4
Developing a plan that delivers significant results in
less time
If there is an obstacle to implementing predictive Retrofit programs typically begin with prioritizing
maintenance, it is equipment connectivity. As a the installed base. Prioritization enables
recent survey notes, up to 92% of legacy machines manufacturers to focus initially on the assets that
are not designed for connectivity. require the most frequent maintenance and have
the biggest impact on production output. After
Manufacturers are understandably reluctant to
critical assets are identified, implementing the
abandon these investments, which is why they
predictive maintenance solution is straightforward.
are turning to retrofitted solutions as a way to
Key steps include replicating engineer knowledge
modernize while keeping costs affordable. Industry
for the targeted asset, establish a baseline of
providers have already brought several technology
acceptable conditions, retrofitting one or more
solutions to market for upgrading legacy assets with
sensors to the asset, and setting warning and
little or no equipment modification.
alarm thresholds.
5
The power of predictive maintenance
Worn components is a leading cause of three- analysis in real time, detect abnormalities, and
phase electric motor failure. Omron recently analyze the failure mode. With this insight, the
helped a beverage company find an effective company has been able to eliminate time-
solution for reducing these failure events. Prior consuming inspections and rely on alerts that
to the solution’s implementation, the company signaled a service requirement. The transition to
manually inspected its bottling line motors at three- a predictive maintenance solution has proven to
and six-month intervals and overhauled motors reduce costs and streamline maintenance work
once a year. With this preventive maintenance while also improving product quality by detecting
strategy, workers replaced parts frequently, but this rubber gasket failures before foreign matter can
approach was expensive and failed to eliminate contaminate products. The company is planning to
unplanned downtime. deploy the solution to other facilities and change
some existing maintenance practices.
Retrofitting the motor with a current transformer
allowed a monitoring AI to run a distortion
6
Use case 2: Recirculating pump condition monitoring
Industrial recirculating pumps run almost continuously. One semiconductor manufacturer used manual
inspections to check a motor in its water treatment plant, but conducting accurate inspections was difficult
without shutting down the pump. Scheduling maintenance was always challenging because of ongoing
production needs. This meant that service might be performed too early (wasting time and materials) or too late
(after an unexpected failure impacted production).
Retrofitting the recirculating pump with a vibration sensor allowed the monitoring AI to measure high-frequency
vibrations, detect abnormalities, and analyze the failure mode. Alerts permit maintenance engineers to monitor
pump health remotely, judge the potential impact of abnormalities, and solve problems without being onsite.
This use case shows how a predictive maintenance solution can streamline maintenance work while also
supporting semiconductor production by preventing unplanned downtime.
7
The power of predictive maintenance
Hydraulic valves are essential to the normal valve temperature, detect abnormalities, and analyze
operation of many industrial machines, and the failure mode. Increases in surface temperature
temperature increases often suggest debris generate automatic alerts that allow maintenance
contamination in the hydraulic fluid flowing through staff to take immediate action if necessary. This
the valves. When a leading auto manufacturer was predictive maintenance solution reduced overall
relying on manual thermal inspections to monitor maintenance needs and improved machine
the condition of valves on its body panel hydraulic availability. Furthermore, since the company no
presses, unplanned downtime lasting days or longer needed to move production to other
weeks sometimes occurred. This was due to the hydraulic presses, efficiency also improved and costs
fact that maintenance engineers were unable to like overtime pay went down.
continuously monitor valve temperature.
8
Use case 4: Power supply with onboard monitoring
Large production environments often house floor. With temperature, voltage, and power sensors,
hundreds of control panels. Temperature increases each power supply is able to calculate its own
in control panel power supply often indicates that remaining service and send a real-time health-
these components are degrading. Maintenance grade to a built-in display. Onboard networking
engineers for one automotive manufacturer also supports remotely monitoring multiple power
were using digital voltmeters to check power supplies from a central location. Using intelligent
supplies for out-of-spec conditions. These manual devices as part of a predictive maintenance
inspections were time-consuming and imprecise solution provides a framework for building an IoT
since components sometimes degraded between environment. It allows intensive monitoring without
inspections. To limit downtime, these workers relied hands-on inspections, improving efficiency and
on preventive schedules to replace power supplies equipment availability while lowering costs.
regardless of device health.
9
The power of predictive maintenance
Summary
10
References
1. Arsenault, Ryan for Aberdeen. (2016). “Stat of the Week: The (Rising!) Cost of Downtime”. Retrieved Friday, October 2, 2020 from
https://www.aberdeen.com/techpro-essentials/stat-of-the-week-the-rising-cost-of-downtime/.
2. Blyler, John. (2018). “Microservices Connect Legacy Assets to the IIoT”. Retrieved October 2, 2020 from https://www.insight.tech/
industry/microservices-connect-legacy-assets-to-the-iiot.
11
Omron Automation | 800.556.6766 | automation.omron.com
U129I-E-01 © 2020 Omron. All Rights Reserved. Sensing | Control | Safety | Vision | Motion | Robotics