Case description:
One manufacturer designates 22 machines as key equipments needing planned maintenance. This company has daily, weekly, monthly and yearly planned maintenance. When asked, their maintenance director said, “We have a very good maintenance system. In the past two years the equipments have never been down by accident.” When asked for relevant data, he said that because the equipments hadn’t been down they had no data.

A senior auditor suggested this company violated TS2 clause 7.5.1.4, saying they didn’t have a predicative maintenance system.

I found I disagree with him, because I think the daily, weekly, monthly and yearly maintenance can be counted as both preventive maintenance and predicative maintenance. I think I might be a little confused with the distinction preventive maintenance and predicative maintenance. May someone help to explain the difference between the two kinds of maintenance?

Case description:
..... I think I might be a little confused with the distinction preventive maintenance and predicative maintenance. May someone help to explain the difference between the two kinds of maintenance?

I believe you mean “Predictive Maintenance”.
See Wikipedia definition Predictive_maintenance
The Overview paragraph provides a good distinction between both approaches.
In short, adding intelligence to an existing preventive maintenance is Predictive maintenance.

Regards,
Govind.

Summary:

PdM or condition based maintenance attempts to evaluate the condition of equipment by performing periodic or continuous (on-line) equipment monitoring. The ultimate goal of PdM is to perform maintenance “just in time”, before the equipment fails in service. This is in contrast to time and/or operation count based maintenance where a piece of equipment gets maintained whether it needs it or not. Time based maintenance is labor intensive, ineffective in identifying problems that develop between scheduled inspections and is not cost effective.

Most PdM inspections are performed while equipment is in service, thereby minimizing disruption of normal system operations. Adoption of PdM in the maintenance of equipment can result in substantial cost savings and higher system reliability.

In the past two years the equipments have never been down by accident.” When asked for relevant data, he said that because the equipments hadn’t been down they had no data.


Another item to be wary of is "have never been down by accident". Some companies play games with downtime, shuffling the books to make it look like the equipment was down for "planned maintenance". You do have to watch out for "the machine just broke, now it is time to do its planned maintenance". I'd ask to see the records for downtime for non-accidental reasons.

Time based maintenance is labor intensive, ineffective in identifying problems that develop between scheduled inspections and is not cost effective.

Or at least, such scheduled maintenance can be ineffective.

I have heard tales (although I haven't seen the raw data) that some operations like changing lightbulbs can be more efficient on a fixed schedule. If you have to run a requisition, call a guy to get a ladder from some other building and then climb up to change one bulb, there is a lot of extra labor. By changing bulbs before they actualy wearout, there is extra material cost, but reduced labor cost. In some cases, this type of maintenance might be the most effective approach.


Tim F

Thank you all. Problem solved.

Also, Predictive maintenance tends to include direct measurement of the item. Example, an infrared picture of a circuit board to determine hot spots.

Preventive tends to be done on time schedules (as you have.) Predictive tends to be direct measurement of the item to determine if it needs maintenance.

Other examples of Predictive Maintenance - evaluation of particles in suspension in a lubricant, sound and vibration analysis of a machine.

Regards,

Dirk

Also, Predictive maintenance tends to include direct measurement of the item. Example, an infrared picture of a circuit board to determine hot spots.

Preventive tends to be done on time schedules (as you have.) Predictive tends to be direct measurement of the item to determine if it needs maintenance.

Other examples of Predictive Maintenance - evaluation of particles in suspension in a lubricant, sound and vibration analysis of a machine.

Regards,

Dirk

I agree with you

Also see: Preventive vs. Predictive Maintenance - What is the difference? TS 16949 - 7.5.1.4. (http://elsmar.com/Forums/showthread.php?t=5000)
and
Preventive vs. Predictive (http://Elsmar.com/Forums/search.php?do=process&titleonly=1&query=preventive+predictive) discussion threads.

:bigwave: In Europe we are involved in a new and Experimental Standards EN (Europeean Norms) for Maintenance.

ENV 13269: Maintenance - Guideline on preparation of maintenance contracts
EN 13306: Maintenance - Terminology
prEN 13460: 2001 Maintenance - Documents for Maintenance
prCEN/TS 15331: Criteria for design, management and control of maintenance services for buildings
prEN 15341: Maintenance - Maintenance Key Performance Indicators

http://www.wear-management.ch/?path=root+standards&lang=en


PDF:

http://www.dienstleistungs-standards.de/englishsite/020903/franlund.pdf

My University of Engineers of Seville (Spain) is leadering these new Standards

None of you guys would make it under my supervision in Army Aviation.

Preventive maintenance was what we performed to detect potential failure (like looking for leaks, wear on rotor blades, checking bearings for wear and so on) and to perform actions to prevent early failure ( cleaning and lubing the swashplate and other bearings, spraypainting leading edges of rotor blades, cleaning/washing engine compressors, polishing windshields, and so on.

Predictive maintenance was essentially maintenace that was planned based on history of the components or airframe such as TBO's (time before overhaul), engine, transmission, blade and flight control replacement, servo rebuilding and things like that.

None of you guys would make it under my supervision in Army Aviation.

Preventive maintenance was what we performed to detect potential failure (like looking for leaks, wear on rotor blades, checking bearings for wear and so on) and to perform actions to prevent early failure ( cleaning and lubing the swashplate and other bearings, spraypainting leading edges of rotor blades, cleaning/washing engine compressors, polishing windshields, and so on.

Predictive maintenance was essentially maintenace that was planned based on history of the components or airframe such as TBO's (time before overhaul), engine, transmission, blade and flight control replacement, servo rebuilding and things like that.This is pretty close to my understanding of difference between preventive and predictive. Things like regular oil changes are predictive; topping off the oil when it is a quart low is preventive.

Pretty much, and it ain't rocket science...by the way they do both preventive and predictive on the shuttle.

I like the earlier example of changing light globes.
A predictive method would probably involve removing and testing the light globe (via some method) and then evaluating the chances of failure before the next test, and replacing the glove if necessary. Or the preventive method, just change it. Obviously there is equipment where significant savings can be made by predicting failures but there is plenty of other equipment where it is not economical to do so.

How does doing oil changes at a pre-determined frequency predict anything?

Doing an oil change is preventive. Analyzing the oil for particles would provide the data to predict a failure.

The data from the maintenance activity is what is used to predict failure before it happens. Direct measurement and data to predict failure modes is the key to predictive. Also, predictive is machine by machine. Stating that all machines of a certain type need a lube job every 6 months is not predictive, it is preventive. Analyzing each machine for specific charactieristics to predict failure is predictive maintenance.

In my experience, the difference between predictive and preventive is not easy for most people. This is exemplified by the existence of this thread, other threads, and the fact that there is disagreemnt amongst us.

Regards,

Dirk

How does doing oil changes at a pre-determined frequency predict anything?

Doing an oil change is preventive. Analyzing the oil for particles would provide the data to predict a failure.

The data from the maintenance activity is what is used to predict failure before it happens. Direct measurement and data to predict failure modes is the key to predictive. Also, predictive is machine by machine. Stating that all machines of a certain type need a lube job every 6 months is not predictive, it is preventive. Analyzing each machine for specific characteristics to predict failure is predictive maintenance.
This is how I've defined them in my mind. I was a bit surprised that some items mentioned here such as oil changes, are considered by some to be predictive. I've always equated preventive maintenance as planned maintenance, typically for a group of objects / machines. Preventive maintenance items may be based on what was at one time predictive measures or data, but preventive maintenance items are items which are known to be necessary based upon long term data.

I somewhat disagree with Randy's saying "Predictive maintenance was essentially maintenace that was planned based on history of the components or airframe such as TBO's (time before overhaul), engine, transmission, blade and flight control replacement, servo rebuilding and things like that." Those items may have been termed predictive in the service years ago, but since they are based upon long term data and they are done, like an oil change, based upon long term data, they are now preventive (done on a routine schedule). I understand Randy's classing an overhaul based upon a TBO as predictive (if you don't do it, the long term data says the engine will fail), but that's the same as an oil change. Long term data indicates that if the oil and filter aren't replaced after a certain amount of time or use (such as miles driven), the engine will fail.

My understanding of predictive is the same as Dirk's "Analyzing the oil for particles would provide the data to predict a failure." example. That is, to use analysis techniques to predict a potential maintenance item or failure which falls outside the field of long term data based maintenance items. That is, it predicts a failure based upon short term or 'snapshot' data. For example, I remember not so long ago doing oil analysis once a year on car motors. High chromium content, for example, typically indicated that the rings were wearing faster than expected so an engine overhaul was done before it was due according to the preventive action plan which was based upon long term data from which the TBO was originally derived.

Now, one may define oil analysis as preventive if it is done every time the oil is changed or on a regular schedule, but it is predictive in that it is an attempt to determine if an unexpected (through consideration of long term data) condition (abnormal ring wear in the example of high chromium) exists. And, although oil analysis may be done on every applicable machine, the results were applicable to each individual machine. You may have two machines that are identical, but the oil analysis of each may predict different scenarios for each machine. Back when I used to fly a lot of people who owned airplanes considered oil analysis as a routine 'preventive' procedure and they called it preventive, but they looked at it from the perspective that it would (theoretically) predict a condition that could not otherwise be identified (particularly engine failure earlier that expected according to existing TBO 'recommendations').

So, as happens in some threads, we have semantics issues popping up. I think the important part is that each person look at their situation and be ready to explain to an auditor what they define, in their specific case, what is predictive and what is preventive. And to be ready to defend their definitions / reasoning.

OK I agree we may be caught in semantics - in effect trying to separate fly specks from pepper.

Marc's suggestion that being able to defend one's semantic choice is important is one of the best in the thread.

I feel confident I could defend complete oil and filter change on a regular schedule (miles or time engine is run) as predictive based on history of events on similar motors, regardless of whether symptoms of failure are discovered during an investigation of any one motor, whereas the topping off of a single quart when the tank is low is reactive. The semantic flurry could then be extended to whether replacing one quart is preventive (keeping an engine from seizing) or "corrective" in terms of correcting a system (engine lubrication) to keep a process running.

I am in agreement with Dirk and Marc. Preventive maintenance is generally carried out to prolong the useful working life of machines and it’s generally based on past experience or manufacturers recommendations.

Wear and tear are two occurrences that nobody can avoid. However well you carried out preventive maintenance, it’s going to fail some day but before that, ample warnings are usually given – in the form of NOISE, VIBRATION and TEMPERATURE. This is where predictive maintenance comes in. Basing again on past experience or techniques such as analysis of oil etc, mechanics and engineers are able to schedule a shut down and repair before a breakdown that affects production schedule (stoppage) happens. Depending on the techniques used, they are often able to predict quite accurately the problems parts and standby replacements so as to minimize downtime.

In the case of electrical powered machines, the increase in power consumption (amp. meter reading) is usually a sign that the motor is working harder than normal – symptom that something is going to give way soon.

I would go further and contend that predictive maintenance is a part of preventive maintenance – actually a technique used in preventive maintenance. All are done to prevent unforeseen shutdowns and maximize the working life of machines.

If you google predictive maintenance, you'll find all kinds of vibration analysis, oil analysis, IR analysis and various testing services. Then google preventive maintenance. That gives a good indication of what the largest proportion of companies consider predictive as opposed to preventive. It appears most link 'Predictive' with 'Diagnostic'.

I tend to agree with how Weibull (http://www.weibull.com/SystemRelWeb/preventive_maintenance.htm) defines Preventive Maintenance. Preventive maintenance is a schedule of planned maintenance actions aimed at the prevention of breakdowns and failures. The primary goal of preventive maintenance is to prevent the failure of equipment before it actually occurs. It is designed to preserve and enhance equipment reliability by replacing worn components before they actually fail. Preventive maintenance activities include equipment checks, partial or complete overhauls at specified periods, oil changes, lubrication and so on. In addition, workers can record equipment deterioration so they know to replace or repair worn parts before they cause system failure. Recent technological advances in tools for inspection and diagnosis have enabled even more accurate and effective equipment maintenance. The ideal preventive maintenance program would prevent all equipment failure before it occurs.

BTW, I sorta jumped in here because back when I did QS-9000 implementations I went through the semantics with lots of auditors in QS-9000 registration audits. Hence my statement about being ready to explain one's definitions and reasoning to an auditor, which is about the only time the distinction is important. In QS-9000 they gave some guidance in 4.9.g.1 "Predictive maintenance methods – These methods should include a review of appropriate items such as the manufacturer’s recommendations, tool wear, optimization of uptime, correlation of SPC data to prevent maintenance activities, important characteristics of perishable tooling, fluid analysis, infrared monitoring of circuits and vibration analysis." None the less, semantics often came into play during audits.

An early related thread here: Preventive and Predictive Maintenance - Information on procedures relating to QS-9000 paragraph 4.9.g.1 (http://elsmar.com/Forums/showthread.php?t=93).

In a simple man's language, I would define the three types of maintenance as follows:

My car goes for a scheduled servicing every quarter / milage basis - PREVENTIVE.

I am rolling on the highway. Suddenly I start getting noise under the hood. I know nothing about how a vehicle runs. I stop at the nearest gas station and tell the mechanic to have a look. - PREDICTIVE.

In the same situation as above, I continue pushing the car till it refuses to run any further and call a tow truck. - BREAKDOWN

Predictive maintenance is basically done when the machine "tells" me that it is going to develop a problem (by itself or because I ask it) and I respond by taking appropriate action - depending on my judgement of the severity of the problem.

In my opinion....

If organizations are doing both and understand the benefits, then I agree it is semantics. But if an organization does not understand the concepts (no matter what they call them), then the difference and the benefits can huge.

The cost of an IR picture of a fuse panel Vs. changing the fuses every 2 years VS replacing the panel because it blew up is significant. Cost and downtime in each case can be very different.

The data gained from predictive maintenance and from experience leads to preventive maintenance. Something has to give us data and a sense of confidence that an engine needs an oil change very 5,000 miles.

Predictive maintenance data is an input into preventive maintenance programs.

Great discussion.

Thank you,

Dirk