PM issues

This may be off subject, but I have a ops manager that is driving us crazy. Were do u draw the line on PM's. He was having use down a line for 2 hours each week to do PM's. What are your opinions on PM"s. I have done a study and in the past 6 months we have only had one major mechanical breakdown that may have been PM'ed out if we had followed this schedule. He has suggested software to monitor this, and has generaly got in the way on several issues pertaining to maintenance. We have a general PM schedule consisting of startup, weekly and monthly checks. I feel this is more than enough. Just curious of what other maintenance departments are doing in this area??

At our plant we have weekly, monthly, bi monthly, quartly, and semi yearly PMs.

All our saws (slicers and bandsaw), dumpers, bagging machines are weekly. Our packaging machines, conveyors and scales are monthly. Depending on the equipment and its usage is how it gets PM'ed. Usually at least once a year just about everything gets torn down once and gone through well (this is usally during summer, our slow time of year).

All of our PM's as well as our Work Orders (regular jobs during each day) are kept via software (Tabware)

PMs done well can be good value.
Examples are greasing barings, changing wear parts and periodic checking of safety systems.
Also, things like barings often have a maximum life, and you can avoid breakdowns if these are changed out a bit before that life is over (assuming they are correctly installed, alligned and lubricated). Also, having a PMed item fail before its time will also indicate a more underlying problem that needs to be looked at.
Contactors and relays can be PMed to change out before their 100,000 operations life ends and they fail.
Some things, like PLCs, don't have much that can be PMed.

Like everything else, you can do PM badly so you get the overhead of the PM and the reliability of breakdown maintenance.

Optimizing PM's can be rewarding, but requires a lot of data tracking.

Here's a simple example:

You perform a PM on widget #1 weekly, you record the findings, unit okay, replaced xyx etc... you also recored everytime widget #1 requires corrective maintenance. Over time you increase the time between PM's from weekly to bi-weekly. You analyze your history, is there a trend of failures, downtime etc..., if not increase to every 3 weeks, analyze again. At some point the failures/downtime will increase, then you reduce your time between PM's so that you catch it just before it occurs.

There are many software packages out there that help to predict and schedule planned maintenance, we have been working at it for several years and still haven't got the results we need/wanted.

I have worked with/co-designed custom PM systems for years with good results and feedback and do have a somewhat strong opinion on this topic. I see it as applied science and take the approach that a properly designed system can provide as much reliability as you desire.

For a good example of what can be achieved just observe any proffesional race on television next chance that you get and see what is actually provided to any team as far as support, research and technology to increase reliability, then look at the results. Of course this is an example of reliability taken to an extremely high level.

This may be off subject, but I have a ops manager that is driving us crazy. Were do u draw the line on PM's. He was having use down a line for 2 hours each week to do PM's.

Click to expand...

You didnt supply any details of the machine/system and its operation. If the machine/system is operational 24/7 and isnt accessible to PM when operating then a weekly PM is the best (maybe only) option. It lets you visually determine areas of concern and take appropriate action. It should be greased, oil checks, bolt checks....etc etc.

There are so many variables involved in PM that the only real answer to your question is "It depends". I have seen many situations where I would liked to have had 2 hours a week to do PM.

As Ron said - it depends! Two hours per week could be overkill OR not nearly enough!

One approach that we have tried (to minimize operational impact) is to increase the downtime hours, but reduce the frequency. So - perhaps four hours every other week, or every third week . . .

Marc

I would LOVE to have a guaranteed 2 hours per week for PM. I have worked in too many facilities where maintenance is always reactive instead of proactive. The lines run 24/7 until something breaks. Then everybody and their cousin is on your back about "how long until you get it up and running again?"
I try to explain that you have 2 choices - give me 2 hours per week when you know when it is coming, or 6-7 hours SURPRISE!!!!!
No one listens.

To emphasize what KBrodin, Doug_Adam, and Ken (in particular) have said - if done right, a good PM program should result in HIGHER uptime on your equipment.


If you are getting LOWER uptime because you are constantly down for PM, then you should consider reducing the frequency (or duration) of your PM shutdowns. If you see your unplanned downtime going up, then you should consider increasing the frequency (or duration) of your PM shutdowns.

Marc


PS: The next level up is "predictive" maintenance (which Doug mentions). Predictive maintenance is when you replace a part BEFORE it will fail based on statistical information about failure rates (a classic example talked about with some regularity on this board is replacing PLC batteries) . . .