Predictive vs. Preventive Maintenance - TS 16949 Clause 7.5.1 - Definitions of

[Stevenli]

Dear all,

how to understand predictive maintenance in TS2 7.5.1? what's the difference with preventive maintenance? and the most important is how to implement it?

Thank you very much
steven

[Marc]

An old post for starters:

Posted by Roger Eastin on Monday, 27 October 1997, at 5:16 p.m., in response to Predictive maintenance methods, posted by Felix cruz on Tuesday, 21 October 1997, at 3:06 p.m.

Predictive maintenance methods mentioned in the QS9 standard are identified by the type of equipment you are operating. If you have machinery that uses items such as gear boxes, then you should consider having vibration analysis done on them (if they are considered key manufacturing equipment). Here the analysis would detect gear wear or possible fractures of the gears. If you have equipment that uses control boards which contain electronic circuits, then you need to look at IR testing of the circuits. This analysis would detect "hot spots" on the circuit which may mean that a component is beginning to break down. If the equipment you are running uses lubricating fluids such as oil, then you would look at using a lab to do fluid analysis. Here you would look for oil breakdown or contamination levels. If your equipment uses perishable tooling such as drill or milling bits, you would set up a program to measure tool wear to anticipate failure. For the first three methods, outside companies can do this. You could probably do this last one. (You could even track MTBF of these parts in order to improve wear by using different vendors, different coatings of bits,etc.)

Originally Posted by Stevenli

Dear all,

how to understand predictive maintenance in TS2 7.5.1? what's the difference with preventive maintenance? and the most important is how to implement it?

Thank you very much
steven

Steven,

"Predictive" seems, to me, another term for "Preventive". If you have a Preventive Maintenance "Process", I wouldn't sweat it. Another standard making its presence known by changing the accepted terminolgy?

[Craig H.]

Originally Posted by energy

Steven,

"Predictive" seems, to me, another term for "Preventive". If you have a Preventive Maintenance "Process", I wouldn't sweat it. Another standard making its presence known by changing the accepted terminolgy?

At the risk of spouting gobbledygook (sp?) here is my take on it.

Predictive maintenance is indeed preventive maintenance, but all preventive maintenance is not predictive. In other words, predictive uses data gathered from the equipment (wear, vibration, etc.) to optimize the interval between instances of a specific maintenance task.

For instance: Why shut down right now to do prev. maint., when predict. maintenance (data gathered) tells us we can go another 100 hours?

Or: It does not matter what the manual says, the oil analysis tells us that the main bearings are toast, and if we don't shut down now, we will be replacing more than bearings... (wham!!! chunka chunka).

JMO

Craig

[db]

Originally Posted by Craig H.

At the risk of spouting gobbledygook (sp?) here is my take on it.

Predictive maintenance is indeed preventive maintenance, but all preventive maintenance is not predictive. In other words, predictive uses data gathered from the equipment (wear, vibration, etc.) to optimize the interval between instances of a specific maintenance task.

For instance: Why shut down right now to do prev. maint., when predict. maintenance (data gathered) tells us we can go another 100 hours?

Or: It does not matter what the manual says, the oil analysis tells us that the main bearings are toast, and if we don't shut down now, we will be replacing more than bearings... (wham!!! chunka chunka).

JMO

Craig

I've had companies that have used MTBF and other measures as part of their predictive maintenance program. Does this fit in as well in your definition Craig?

[Craig H.]

Originally Posted by db

I've had companies that have used MTBF and other measures as part of their predictive maintenance program. Does this fit in as well in your definition Craig?

Without a doubt...

Originally Posted by Craig H.

For instance: Why shut down right now to do prev. maint., when predict. maintenance (data gathered) tells us we can go another 100 hours?

Craig

Let's take my car as an example. The manufacturer requires that I change the oil every 3K miles. Should I analyze the oil and based on the results, extend the intervals an extra 1 or 2 K miles? Do I know more than the equipment manufacturer? And, do I spend big bucks on analysis vs. following the Manufacturer's recommendations. I don't know. If the Customer requires it, charge them more per unit?

[db]

Originally Posted by energy

Let's take my car as an example. The manufacturer requires that I change the oil every 3K miles. Should I analyze the oil and based on the results, extend the intervals an extra 1 or 2 K miles? Do I know more than the equipment manufacturer? And, do I spend big bucks on analysis vs. following the Manufacturer's recommendations. I don't know. If the Customer requires it, charge them more per unit?

My thinking:

If you change your oil, based on the manufacturer's requirements, then it would be preventive

If you change your oil, based on the analsis of the oil then it would be predictive.

The question is should we really worry about the difference?