Force it on, latch it out, or shut down and fix it??

So I am wondering how many other troubleshooters force things on or write around things temporarily to get a machine running, then come back and fix the problem.

I spend a large part of my time role playing situations so I am prepared when something goes wrong. I now can write around most of our problems rather than shut down to change a prox or adjust a photo eye, I just write around it for a min till lunch. Knowing the machine well has a big part in this. Don't force or write around pressure switches or safety devices or stuff you don't know about obviously.

In our shop so far it seams the people saying fix it, or that things should not be written around spend their time ****ing off while the ones who agree with me seem to spend their time understanding the machines and reading all the tech manuals. ( side note the naysayers are old and think computers are dumb)

The question
Accrue 5 min of downtime to latch it out/ or force it on?

Or an hour to fix it?

Years ago when I was an maintenance engineer, we had many sensors linked out, e-stops disconnected, safety switched bypassed around the whole factory. It was an accident waiting to happen.

In that place: production > everything else.

My job now is just programming and some commissioning as i'm an SI these days. I only deal with new projects and not old machines, sure I have my commissioning bits when on-site but they are never left on.

Years after I left, the factory I used to work in got fined for breeches of H&S Law, they got caught on more than 1 occasion with safety systems linked out.

I suspect the attitude has changed from the 'just get it going' now they had a big fine to pay.

MitsM83 said:

( side note the naysayers are old and think computers are dumb)

Click to expand...

I could not help but think.....and just had to ask.

These "OLD Naysayers", just what age would you say they are on average?

BCS

If you know the machine well enough to, in five minutes:
Identify the problem
Redo the risk assessment
Assess the performance of the machine without the sensor
Make a decision
Enable the force
Retrain the operator for any new operation requirements
Put in place something to make sure the bandaid doesn't become permanent.

Then yes, what you are talking about is going to save the company money.

I think it is also true that your colleagues may find the above list will take longer for them. Plus if they get to the decision making stage and they go the other way, they will then have the downtime increased by the risk assessment/performance time.

It sounds like you know a lot about your machines, and there are a lot of sensors for closed loop control that don't effect safety or quality. Maybe work out a program to automatically (or manually) enable a 'limp' mode, and raise a flag that maintenance personnel are required to service this machine at the next downtime.

Also, an hour of downtime to replace/realign a prox? How old are YOU?

janner_10 said:

Years ago when I was an maintenance engineer, we had many sensors linked out, e-stops disconnected, safety switched bypassed around the whole factory. It was an accident waiting to happen.

In that place: production > everything else.

My job now is just programming and some commissioning as i'm an SI these days. I only deal with new projects and not old machines, sure I have my commissioning bits when on-site but they are never left on.

Years after I left, the factory I used to work in got fined for breeches of H&S Law, they got caught on more than 1 occasion with safety systems linked out.

I suspect the attitude has changed from the 'just get it going' now they had a big fine to pay.

Click to expand...

I agree that is a problem. We have separate plc for safety devises and if it's in there I don't touch it.
We have a auto backup system that runs a verify once a week and emails all of us the changes. It's also a rule to put it in our mms and don't leave anything across shifts without physically showing the in comming people.

Returning to fix the issue is very important. Temp fixes can never be permanent solutions.

The old folks I mentioned are a combined average of 59. Don't get me wrong those guys have forgot more than I know! We have some older guys that are on it. They study and stay up with tech. The two main ones I was referring to there can't even check their email. How they still work here is beyond me.

So what happened is a shift was left without a nerd for a week and it hurt production. These nonnerds are saying never ever latch or force. I agree with them that they shouldn't that could be dangerous but the ones who can should if that makes sense.

Production allways trumps everything in my industry.

AustralIan said:

If you know the machine well enough to, in five minutes:
Identify the problem
Redo the risk assessment
Assess the performance of the machine without the sensor
Make a decision
Enable the force
Retrain the operator for any new operation requirements
Put in place something to make sure the bandaid doesn't become permanent.

Then yes, what you are talking about is going to save the company money.

I think it is also true that your colleagues may find the above list will take longer for them. Plus if they get to the decision making stage and they go the other way, they will then have the downtime increased by the risk assessment/performance time.

It sounds like you know a lot about your machines, and there are a lot of sensors for closed loop control that don't effect safety or quality. Maybe work out a program to automatically (or manually) enable a 'limp' mode, and raise a flag that maintenance personnel are required to service this machine at the next downtime.

Also, an hour of downtime to replace/realign a prox? How old are YOU?

Click to expand...

I try to make a HMI button for anything that can safely be done with limited understanding. I'm working on a rfid badge security system to protect the hmi screens, so far passwords have not kept the operators out. Once I know only mx can get on the mx screens I'll add buttons for anything I do more than twice.
I'm old enough to know I don't know everything but young enough to think I can learn it haha.

While there are times and places for "patching" to get or keep things running, remember "There's nothing so permanent as Temporary".

^Exactly. A large part of my business is rewriting old programs from the past that have been maintained over the years by plant technicians. I spend weeks identify code that's no longer in service, sensors that no longer work, I/O in the code that's been AFI or jumper out. Most times wr spend time fixing all the devices from years of neglect and thus removing all the patches. I am not in favor of patching or temporary fixes. Fix it right the first time.

Personally I don't force anything - if an input needs to be forced on I jumper the input or if something needs disabled I remove the wire from it.

When I go online with a PLC and see the Forces On lit I get a knot in my stomach. Then I do what I can to remove forces. If a line needs disabled I put in an AFI or if an instruction needs to be true I branch around with a XIC Always On bit and remove forces

IF you can put a "workaround" into the code that is 100% safe, then it could have been done that way in the first place ???

Editing any code "to make it produce widgets" is always going to be the driving force from a commercial standpoint, but that approach will, in almost every scenario, make the application code less "secure", less fault tolerant, and inevitably, less "safe".

The PLC program is not "broken" when Auto-Valve AV347 loses a feedback sensor, so why fix it in the PLC... You repair the problem, not sweep it under the carpet..

my 2c

In an ideal world you would lock out the machine, send the operator to your spare one and fix the root Problem. I once had a customer that would keep whole sub-assemblies in stock. If a repair was going to take more then 10 minutes the defect sub-assembly would be brought to the workshop and completely overhauled, costly but consequent.
But sometimes there will
- be a foreman that needs exactly this machine running now,
- be an empty storage bin where the part you need to fix the machine should be according to your inventory system,
- none of your maintanence crew available. Management figured that there would be no machine to fix for them with a machine availability goal of 100%, which led them to cut down on manning...

You will have to bodge something together, defy physics in cases of mechanical faults but NEVER EVER bypass or force safety devices. Out of my own experience I can tell you that it ain't funny to open some safety door and see all axis of a handling system still moving. I then learned that the switch was jumpered out because the door would swing open occasionally...

Keep the boomerang effect in mind, most of the times all of the creativity you put into hiding symptoms will bite you back when the underlying problem is fixed during the machines PM for this decade. Always document your changes, make it easy for the guy that fixes the problem to revert to the standard program. Delete all of the bypass logic immediately after that or it will become a functionality of the machine and you end up adding a button to the HMI for it...

If you are working at a nuclear plant, forget all of the above

Daba,

I didn't mean a broken sensor, I was thinking more like something was removed from the machine (or added) years back and the input or output was forced instead of making a wiring or program change.

I just don't like to see that forced indicator on - and when I go online with a MicroLogic and it always shows the Forces Enabled ON it takes me a few seconds to remember it's normal for them.

I tend to agree that problems should not be solved on the PLC as that leads to the maintenance crew essentially becoming an operator and not putting enough pressure on management to fix whatever is broken by consistently forcing control enigneers to just overwrite it in code.

However, I've been in situations where by not doing this on the PLC, someone else will be put in a far more dangerous situation for operation to go on.
Where I worked, a guy would move into a machine and actuate the hydraulic leavers to move a machine next to a structure running the risk of getting crushed if something went tits up for some reason.
So until I traced the ground fault that caused the issue with the positioning, any of the operators could take me out of bed (this was in a ship) to move the machine if needed, no questions asked. The machine itself was very very poor in terms of functionality.

Another was when in the space of 2 hours 4 out of 6 P+F proximity switches died. We had 3, the current status of operation would not be safe to stop and replace at least 2 of them, so I changed the polarity of the signals in the Ex barrier to keep the system happy. This was accompanied by a talk with the operators, a signed risk assessment and no operation time lost. However, I had 6 years experience in the control system development of those machines and know it inside out.

There are a few old saying I shall bring up.
"Don't fix hardware problems with software" i.e. faulty sensor.
"Fix it right the first time", if you patch something the operators will think is fixed, or you might forget to fix it later. I hope you are not bypassing safety sensors.

I work for a big company as an automation tech. I know there are many customers with this mentality, patch it quickly in the software, bypass a limit, etc. Without bringing up all the paperwork I have to do to make a code change, NOT parameter change, my ethics won't allow me to patch things up just for the sake of keeping operations running. Proper maintenance is already not done properly in most plants so you are not really helping to the cause by using quick patches.

I would recommend a few engineering ethic books where one minor issue cascaded into more complex catastrophic issue in some occasions even deadly, but I can't remember the title right now.

So in short, I never force any outputs. I find the reason why the output is not firing up, and then inform the customer what is part is faulty or whatever needs adjustment.

Thanks for the input guys.
As I said never ever force or latch out... or do anything to a safety device. I would not even touch a output that moves something, like a Drive output be careful with limit switches too.

No matter how much money you save the company, it is not worth a life. That is what you have to keep in mind.

That said if it absolutely can not hurt anyone and it saves a million or so an hour of downtime....Do it.

I am over here in awe that some of you have found E-Stops or safety stuff gone around.... Some people are Crazy!