Department of Industrial Engineering

Special Topics in Industrial Engineering

Predictive Maintenance

Prepared by: Azamat Kibekbaev

Student No: 50071102

Supervised by: Prof. Dr. CEVDET MERIC

December 2011

Abstract

This paper aims at proposing reference models to integrate the predictive maintenance in

manufacturing system. To keep our operations running on schedule and budget, predictive

maintenance management has become an important part of our overall business strategy. We

need to proactively manage our predictive maintenance activities efficiently and effectively by

increasing collaboration between our production and maintenance resources, while optimizing

the performance of our equipment. Maintenance costs are a major part of the total operating costs

of all manufacturing or production plants. Depending on the specific industry, maintenance costs

can represent between 15 and 60 percent of the cost of goods produced. For example, in food

related industries, average maintenance costs represent about 15 percent of the cost of goods

produced, whereas maintenance costs for iron and steel, pulp and paper, and other heavy

industries represent up to 60 percent of the total production costs.

Introduction

What is maintenance and why is it performed? Past and current maintenance practices in both the

private and government sectors would imply that maintenance is the actions associated with

equipment repair after it is broken. The dictionary defines maintenance as follows: “the work of

keeping something in proper condition; upkeep.” This would imply that maintenance should be

actions taken to prevent a device or component from failing or to repair normal equipment

degradation experienced with the operation of the device to keep it in proper working order.

Unfortunately, data obtained in many studies over the past decade indicates that most private and

government facilities do not expend the necessary resources to maintain equipment in proper

working order. Rather, they wait for equipment failure to occur and then take whatever actions

are necessary to repair or replace the equipment. Nothing lasts forever and all equipment has

associated with it some predefined life expectancy or operational life. For example, equipment

may be designed to operate at full design load for 5,000 hours and may be designed to go

through 15,000 starts and stop cycles (1).

The need for maintenance is predicated on actual or impending failure – ideally, maintenance is

performed to keep equipment and systems running efficiently for at least design life of the

component(s). As such, the practical operation of a component is time-based function. If one

were to graph the failure rate a component population versus time, it is likely the graph would

take the “bathtub” shape shown in Figure 5.1.1. In the figure the Y axis represents the failure rate

and the X axis is time. From its shape, the curve can be divided into three distinct: infant

mortality, useful life, and wear-out periods.

The initial infant mortality period of bathtub curve is characterized by high failure rate followed

by a period of decreasing failure. Many of the failures associated with this region are linked to

poor design, poor installation, or misapplication. The infant mortality period is followed by a

nearly constant failure rate period known as useful life. There are many theories on why

components fail in this region, most acknowledge that poor O&M often plays significant role. It

is also generally agreed that exceptional maintenance practices encompassing preventive and

predictive elements can extend this period. The wear-out period is characterized by a rapid

increasing failure rate with time. In most cases this period encompasses the normal distribution

of design life failures (2).

The design life of most equipment requires periodic maintenance. Belts need adjustment,

alignment needs to be maintained, and proper lubrication on rotating equipment is required, and

so on. In some cases, certain components need replacement, (e.g., a wheel bearing on a motor

vehicle) to ensure the main piece of equipment (in this case a car) last for its design life. Anytime

we fail to perform maintenance activities intended by the equipment’s designer, we shorten the

operating life of the equipment. But what options do we have? Over the last 30 years, different

approaches to how maintenance can be performed to ensure equipment reaches or exceeds its

design life have been developed in the United States. In addition to waiting for a piece of

equipment to fail (reactive maintenance), we can utilize preventive maintenance, predictive

maintenance, or reliability centered maintenance (2).

(Lindley R. Higgins, Dale P. Brautigam, and R. Keith Mobley, Maintenance Engineering)

Literature Review

Reactive Maintenance

Reactive maintenance is basically the “run it till it breaks” maintenance mode. No actions or

efforts are taken to maintain the equipment as the designer originally intended to ensure design

life is reached. Studies as recent as the winter of 2000 indicate this is still the predominant mode

of maintenance in the United States. The referenced study breaks down the average maintenance

program as follows:

55% Reactive

31% Preventive

12% Predictive

2% Other

Note that more than 55% of maintenance resources and activities of an average facility are still

reactive.

Advantages to reactive maintenance can be viewed as a double-edged sword. If we are dealing

with new equipment, we can expect minimal incidents of failure. If our maintenance program is

purely reactive, we will not expend manpower dollars or incur capital cost until something

breaks. Since we do not see any associated maintenance cost, we could view this period as

saving money. The downside is reality. In reality, during the time we believe we are saving

maintenance and capital cost, we are really spending more dollars than we would have under a

different maintenance approach. We are spending more dollars associated with capital cost

because, while waiting for the equipment to break, we are shortening the life of the equipment

resulting in more frequent replacement. We may incur cost upon failure of the primary device

associated with its failure causing the failure of a secondary device. This is an increased cost we

would not have experienced if our maintenance program was more proactive. Our labor cost

associated with repair will probably be higher than normal because the failure will most likely

require more extensive repairs than would have been required if the piece of equipment had not

been run to failure. Chances are the piece of equipment will fail during off hours or close to the

end of the normal workday. If it is a critical piece of equipment that needs to be back on-line

quickly, we will have to pay maintenance overtime cost. Since we expect to run equipment to

failure, we will require a large material inventory of repair parts. This is a cost we could

minimize under a different maintenance strategy (2).

Preventive Maintenance

Preventive maintenance can be defined as follows: Actions performed on a time- or machine-

run-based schedule that detect, preclude, or mitigate degradation of a component or system with

the aim of sustaining or extending its useful life through controlling degradation to an acceptable

level.

The U.S. Navy pioneered preventive maintenance as a means to increase the reliability of their

vessels. By simply expending the necessary resources to conduct maintenance activities intended

by the equipment designer, equipment life is extended and its reliability is increased. In addition

to an increase in reliability, dollars are saved over that of a program just using reactive

maintenance. Studies indicate that this savings can amount to as much as 12% to 18% on the

average. Depending on the facilities current maintenance practices, present equipment reliability,

and facility downtime, there is little doubt that many facilities purely reliant on reactive

maintenance could save much more than 18% by instituting a proper preventive maintenance

program.

While preventive maintenance is not the optimum maintenance program, it does have several

advantages over that of a purely reactive program. By performing the preventive maintenance as

the equipment designer envisioned, we will extend the life of the equipment closer to design.

This translates into dollar savings. Preventive maintenance (lubrication, filter change, etc.) will

generally run the equipment more efficiently resulting in dollar savings. While we will not

prevent equipment catastrophic failures, we will decrease the number of failures. Minimizing

failures translate into maintenance and capital cost savings (8).

Reliability Centered Maintenance

Reliability centered maintenance (RCM) magazine provides the following definition of RCM: “a

process used to determine the maintenance requirements of any physical asset in its operating

context.”

Basically, RCM methodology deals with some key issues not dealt with by other maintenance

programs. It recognizes that all equipment in a facility is not of equal importance to either the

process or facility safety. It recognizes that equipment design and operation differs and that

different equipment will have a higher probability to undergo failures from different degradation

mechanisms than others. It also approaches the structuring of a maintenance program

recognizing that a facility does not have unlimited financial and personnel resources and that the

use of both need to be prioritized and optimized. In a nutshell, RCM is a systematic approach to

evaluate a facility’s equipment and resources to best mate the two and result in a high degree of

facility reliability and cost-effectiveness. RCM is highly reliant on predictive maintenance but

also recognizes that maintenance activities on equipment that is inexpensive and unimportant to

facility reliability may best be left to a reactive maintenance approach. The following

maintenance program breakdowns of continually top-performing facilities would echo the RCM

approach to utilize all available maintenance approaches with the predominant methodology

being predictive.

Because RCM is so heavily weighted in utilization of predictive maintenance technologies, its

program advantages and disadvantages mirror those of predictive maintenance. In addition to

these advantages, RCM will allow a facility to more closely match resources to need while

improving reliability and decreasing cost (3), (9).

Predictive Maintenance

Predictive maintenance can be defined as measurements that detect the onset of degradation

mechanism, thereby allowing causal stressors to be eliminated or controlled prior to any

significant deterioration in the component physical state. Results indicate current and future

functional capability. Basically, predictive maintenance differs from preventive maintenance by

basing maintenance need on the actual condition of the machine rather than on some preset

schedule. Preventive maintenance is time-based. Activities such as changing lubricant are based

on time, like calendar time or equipment run time. For example, most people change the oil in

their vehicles every 5,000 to 8,000 Kms travelled. This is effectively basing the oil change needs

on equipment run time. No concern is given to the actual condition and performance capability

of the oil. It is changed because it is time. This methodology would be analogous to a preventive

maintenance task. If, on the other hand, the operator of the car discounted the vehicle run

time and had the oil analyzed at some periodicity to determine its actual condition and

lubrication properties, he/she may be able to extend the oil change until the vehicle had

travelled 15,000 Kms. This is the fundamental difference between predictive maintenance and

preventive maintenance, whereby predictive maintenance is used to define needed maintenance

task based on quantified material/equipment condition.

Predictive maintenance can be defined as follows also: Measurements that detect the onset of

system degradation (lower functional state), thereby allowing causal stressors to be eliminated or

controlled prior to any significant deterioration in the component physical state. Results indicate

current and future functional capability.

Basically, predictive maintenance differs from preventive maintenance by basing maintenance

need on the actual condition of the machine rather than on some preset schedule. You will recall

that preventive maintenance is time-based. Activities such as changing lubricant are based on

time, like calendar time or equipment run time. For example, most people change the oil in their

vehicles every 3,000 to 5,000 miles traveled. This is effectively basing the oil change needs on

equipment run time. No concern is given to the actual condition and performance capability of

the oil. It is changed because it is time. This methodology would be analogous to a preventive

maintenance task. If, on the other hand, the operator of the car discounted the vehicle run time

and had the oil analyzed at some periodicity to determine its actual condition and lubrication

properties, he/she may be able to extend the oil change until the vehicle had traveled 10,000

miles. This is the fundamental difference between predictive maintenance and preventive

maintenance, whereby predictive maintenance is used to define needed maintenance task based

on quantified material/equipment condition.

The advantages of predictive maintenance are many. A well-orchestrated predictive

maintenance program will all but eliminate catastrophic equipment failures. We will be able to

schedule maintenance activities to minimize or delete overtime cost. We will be able to minimize

inventory and order parts, as required, well ahead of time to support the downstream

maintenance needs. We can optimize the operation of the equipment, saving energy cost and

increasing plant reliability. Past studies have estimated that a properly functioning predictive

maintenance program can provide a savings of 8% to 12% over a program utilizing preventive

maintenance alone. Depending on a facility’s reliance on reactive maintenance and material

condition, it could easily recognize savings opportunities exceeding 30% to 40%. In fact,

independent surveys indicate the following industrial average savings resultant from initiation of

a functional predictive maintenance program:

Return on investment: 10 times

Reduction in maintenance costs: 25% to 30%

Elimination of breakdowns: 70% to 75%

Reduction in downtime: 35% to 45%

Increase in production: 20% to 25%.

On the down side, too initially start into the predictive maintenance world is not inexpensive.

Much of the equipment requires cost in excess of $50,000. Training of in-plant personnel to

effectively utilize predictive maintenance technologies will require considerable funding.

Program development will require an understanding of predictive maintenance and a firm

commitment to make the program work by all facility organizations and management.

Predictive maintenance is a condition-driven preventive maintenance program. Instead of relying

on industrial or in-plant average-life statistics, i.e. mean-time-to-failure, to schedule maintenance

activities, predictive maintenance uses direct monitoring of the operating condition, efficiency,

heat distribution and other indicators to determine the actual mean-time-to-failure or loss of

efficiency that would be detrimental to plant operations for all critical systems in the plant or

facility. At best, traditional time-driven methods provide a guideline to normal machine-train life

spans. The final decision, in preventive or run-to-failure programs, on repair or rebuild schedules

must be made on the bases of intuition and the personal experience of the maintenance manager.

The addition of a comprehensive predictive maintenance program can and will provide factual

data on the actual operating condition of critical assets, including their efficiency, as well as the

actual mechanical condition of each machine- train and the operating efficiency of each process

system. Instead of relying on industrial or in-plant average-life statistics, i.e. mean-time-to-

failure, to schedule maintenance activities, predictive maintenance uses direct monitoring of the

mechanical condition, system efficiency and other indicators to determine the actual mean-time-

to-failure or loss of efficiency for each machine-train and system in the plant. This data provides

maintenance management the factual data needed for effective planning and scheduling

maintenance activities.

Predictive maintenance is much more. It is the means of improving productivity, product quality

and overall effectiveness of our manufacturing and production plants. Predictive maintenance is

not vibration monitoring or thermal imaging or lubricating oil analysis or any of the other

nondestructive testing techniques that are being marketed as predictive maintenance tools.

Rather, it is a philosophy or attitude that simply stated uses the actual operating condition of

plant equipment and systems to optimize total plant operation. A comprehensive predictive

maintenance management program utilizes a combination of the most cost-effective tools, i.e.

thermal imaging, vibration monitoring, tribology and other nondestructive testing methods, to

obtain the actual operating condition of critical plant systems and based on this factual data

schedules all maintenance activities on an as-needed basis.

Including predictive maintenance in a comprehensive maintenance management program will

provide the ability to optimize the availability of process machinery and greatly reduce the cost

of maintenance. It will also provide the means to improve product quality, productivity and

profitability.

A predictive maintenance program can minimize unscheduled breakdowns of all electrical and

mechanical equipment in the plant and ensure that repaired equipment is inacceptable condition.

The program can also identify problems before they become serious. Most problems can be

minimized if they are detected and repaired early. Normal mechanical failure modes degrade at a

speed directly proportional to their severity. If the problem is detected early, major repairs can

be prevented, in most instances (5), (6).

Benefits (4)

Effective use of preventive maintenance, including predictive technologies, will eliminate much

of the 33 % to 50 % of maintenance expenditures that are wasted by most manufacturing and

production plants. Based on historical data in the USA, the initial savings generated by effective

preventive/predictive maintenance programs fall into the following areas:

1. Elimination of unscheduled downtime caused by equipment or system failures. Typically,

reductions of 40 % to 60 % are achieved within the first two years and up to 90 % reductions

have been achieved and sustained within five years.

2. Increased manpower utilization. Statistically, the average “wrench-time” of a maintenance

craftsperson is 24.5 % or about 2 hours per shift. By identifying the precise repair task needed to

correct deficiencies within a plant asset, as well as the parts, tools and support needed to rectify

the problem, preventive/predictive maintenance can dramatically increase effective “wrench-

time”. Most plants have been able to achieve and sustain 75 % to 85 % effective utilization.

3. Increased capacity. The primary benefit of effective preventive/predictive maintenance

programs is an increase in the throughput or production capacity of the plant. Short-term, i.e. 1-

to-3 years, increases in sustainable capacity have ranged between 15 % and 40 %. Long-term

improvements of 75 % to 80 % have been achieved.

4. Reduction of maintenance expenditures. In some cases, actual maintenance expenditures will

increase during the first year following implementation of an effective preventive/ predictive

program. This increase, typically 10 % to 15 %, is caused by the inherent reliability problems

discovered by the use of predictive technologies. When these problems are eliminated, the

typical result is a reduction in labor and material cost of between 35 % and 60 %.

5. Increased useful life. Typically, the useful operating life of plant assets will be extended by 33

% to 60 %. Detecting incipient problems or deviations from optimum operating conditions

before damage to equipment occurs derives this benefit. Making minor adjustments or repairs

and not permitting a minor deficiency from becoming a serious problem can extend the effective

useful life extended almost indefinitely.

Predictive maintenance, on the other hand, determines when the machine REQUIRES repair.

Plant machinery is therefore only repaired WHEN REQUIRED. The benefits of predictive

maintenance can be separated into two main categories.

INCREASED SAFETY: Predictive maintenance provides the reassurance of safe, continued

plant operation. By reducing the likelihood of unexpected equipment breakdown, the safety of

employees is improved. Although difficult to quantify, there is a definite economic benefit in

improved employee and union relationships.

IMPROVED OPERATING EFFICIENCY: There are many areas in which a predictive

maintenance program can increase the efficiency of your process. Please see chart below.

(SACHS, Neville W., Predictive Maintenance Cuts Costs)

Predictive maintenance requirements (4)

Predictive maintenance is mainly based on the study of the variations of the system status in

terms of pertinent information or data materializing the degradation and the drift behavior of the

system.

The dynamic of degradation describes its past, present and future values, as a control variable, to

represent the degradation through time (Figure 2). This representation is required to implement

function linked to the predictive maintenance objectives.

In this way, predictive maintenance has to control degradation by acting on the degraded system

to realise a nominal or a degraded function. This action allows applying resumption, emergency

procedure, reconfiguration, key locking, redundant functioning…

So, the concept of predictive maintenance is closed to control theory, but by considering that the

predictive maintenance goal is to monitor, diagnose, and prognosticate online the degradation of

manufacturing system to compensate, correct or operate just in time.

To act on the degraded manufacturing system, the predictive maintenance must authorize actions

only in total coherence with the manufacturing system behavior in order to be sure that these

actions allow retrieving a normal state from an abnormal one in safety conditions.

Predictive maintenance on the production line (7)

Nowadays maintenance is considered as a key point for the manufacturing system

competitiveness because first its cost represents the major part of the operational cost, and

second, a system failure can have an important impact on product quality, equipment availability,

environment, and operator.

The two reasons for stopping a production line are due to regularly scheduled maintenance or

equipment failure. Performing timely maintenance is critical to preventing failures that may

result in costly production interruptions, but relying on a fixed schedule may result in higher than

necessary costs for both parts and labor. Predictive maintenance leverages the rich set of data

those manufacturers already have available, such as equipment type, number of days in

operation, operating voltage, days from last service, days to next service, failure history, costs for

planned and unplanned maintenance, parts analysis and other data depending upon the

machinery involved.

A fully automated process analyzes this data in real time. It quickly detects failure patterns and

identifies the root cause of the problem. Because engineers have 24/7 access to the reliability of

every piece of equipment, they can evaluate the current status of every asset and build a

maintenance schedule that performs inspections and/or maintenance just in time to prevent

failures. This eliminates the need for shutting down a line simply to perform “regularly

scheduled maintenance” that may not be really necessary.

As operating conditions change, the reliability of every piece of equipment is updated in real

time. The advanced algorithms contained in the predictive maintenance software can determine

the reliability of every asset at any point in the future, so that inspections and maintenance can be

performed at the optimal and most cost-effective moment. Predictive maintenance also identifies

the replacement parts required to support this highly accurate maintenance schedule. It

eliminates the need for unnecessary and expensive overstocking of spare parts.

Using predictive maintenance, the manufacturer can determine if certain production runs fail

more often than others. They conduct a root cause analysis to determine the problem’s source,

and then analyze the financial implications of the failure to determine whether those runs warrant

a recall or just a service bulletin to their distributors. Their analysis also shows where the failures

will occur and what the demand in a given region will be for the replacement parts. They can

then ensure that the correct supply of replacement parts can be available at the appropriate time.

Reducing warranty claims (7)

As a result, the company avoids many costly warranty claims by providing the resolution to its

service channel even before most customers know a problem exists.

In many situations like this, predictive maintenance can identify when equipment in the field is

likely to fail or need maintenance in order to predict future warranty claims costs and maximize

uptime/in-service time for equipment sold to customers or for equipment used to deliver service.

This helps manufacturers avoid high services costs and product recalls due to late product issue

identification. It also minimizes or eliminates bad publicity and the resulting lost sales from

recalls or negative customer product reviews.

Advantages and Disadvantages of Predictive Maintenance

Predictive maintenance is a more condition-based approach to maintenance. The approach is

based on measuring of the equipment condition in order to assess whether equipment will fail

during some future period, and then taking action to avoid the consequences of that failures. This

is where predictive technologies (i.e. vibration analysis, infrared thermographs, ultrasonic

detection, etc.) are utilized to determine the condition of equipment, and to decide on any

necessary repairs. This approach is more economically feasible strategy as labors, materials and

production schedules are used much more efficiently.

Advantages of Predictive Maintenance

Provides increased component operational life and availability.

Results in decrease in equipment and/or process downtime.

Lowers costs for parts and labor.

Provides better product quality.

Improves worker and environmental safety

Raises worker morale.

Increases energy savings.

Increased component operational life/availability

Allows for preemptive corrective actions

Decrease in equipment or process downtime

Better product quality

Improved worker and environmental safety

Improved worker morale

Energy savings.

Estimated 8% to 12% cost savings over preventive maintenance programme

Disadvantages of Predictive Maintenance

Increases investment in diagnostic equipment.

Increases investment in staff training.

Savings potential is readily seen by management.

Developing a predictive maintenance application

While each company is different, a typical approach to developing a predictive maintenance

application can be found below:

1. Identify the problem by focusing on failures of specific asset types or specific events.

2. Notice as many data sources (e.g. asset information, maintenance history, inspection reports,

RFIDs) as possible.

3. Integrate all data sources, such as combining “fixed” attributes (e.g. asset type) and “dynamic”

attributes (e.g. temperature).

4. Derive any additional fields to help with modeling, such as creating a “1” or “0” to signify if

a part failed or not or calculating average cost per part, in addition to existing data of total cost

and number of parts.

5. Identify the best predictors of failure.

6. Evaluate resultant models for modeling accuracy and quick logic test based on your own

experiences.

7. Focus on the most effective methods of deployment within your organization (e.g. asset

management, workforce management systems, business intelligence).

8. Create an information and response feedback loop to maintain accuracy and make continual

improvements.

9. Monitor and track progress.

Conclusion

As manufacturers face increasing pressures to control costs and improve productivity, preventive

maintenance has emerged as an essential capability. Supported by predictive analytics, predictive

maintenance prevents production interruptions, improves usability and service levels for

customers and helps to reduce warranty costs. It empowers manufacturers to isolate and solve

maintenance and operational issues before they become significant and expensive problems. For

manufacturers that need to achieve the highest levels of uptime and minimize the expense and

labor of downtime, it’s a game-changing technology.

Artificially high maintenance costs caused by a combination of ineffective management methods

and the lack of timely, factual knowledge of asset condition represent a substantial opportunity

for almost every manufacturing and production facility worldwide. Effective use of the

preventive/predictive technologies provides the means to take advantage of this opportunity.

Used correctly, the 33 % to 50 % of wasted maintenance expenditures can be eliminated and

effective use of plant resources; both production and maintenance can be achieved and sustained.

Acknowledgment:

I want to thank our professor CEVDET MERIC for helping and supporting us in this course and

give us the chance to prepare this project.

References

1) Lindley R. Higgins, Dale P. Brautigam, and R. Keith Mobley, Maintenance Engineering

Handbook, McGraw Hill Text, 5th Edition, September 1994

2) Joseph D. Patton, Jr. Maintainability and Maintenance Management, Instrument Society

of America, 3rd Revision, February 1994

3) John Moubray, Reliability-Centered Maintenance, Industrial Press, 2nd Edition, April

1997

4) J-b. Leger, E. Neunreuther, B. Iung, G. Morel, cran-gsip Nancy Research Centre of

Automatic Control. “Integration of the Predictive Maintenance in Manufacturing

System”, 2010

5) Mobley K, Introduction to Predictive Maintenance. Van Nostrand Reinhold, New York,

1989

6) Mobley K, Predictive Maintenance Handbook , 1998

7) ibm.com/software/analytics/industrial-products. Predictive Maintenance for

Manufacturing How to improve productivity, minimize downtime and reduce costs.2011

8) Moinuddin Madki, Workshop on: Predictive & Preventive Maintenance Basics &

Practices. 19 – 20, September 2011

9) NASA. Reliability Centered Maintenance Guide for Facilities and Collateral

Equipment. National Aeronautics and Space Administration, Washington, D.C. 2000

10) SACHS, Neville W., Predictive Maintenance Cuts Costs , Power Transmission Design ,

November 1986.