So I just started a new job at a plant that makes seals, similar to pic related. Even though I'm new, I've been asked to sit in on a meeting regarding an issue they've had, and I think it would be nice if I had some input.
As for the issue...the production crew has recently gotten into the bad habit of mixing up seals. For exampl.e, one seal might have be 4 x 5 (ID x OD), and it'll get dumped in with a box of seals that are 100mm x 125mm. Very close but not the same. Now we're trying to find a fast and simple way for the monkeys in production to differentiate things.
Off the top of my head, I've thought of some kind of go/no go gauge, or some of visual marker or stamp. Any other ideas?
try this maybe?
>>944095
or something like it, easy to differentiate patterns?
>>944090
> mixing metric with imperial
> why
Put a feed pipe on the part hopper that is small enough to allow the smaller part through, but too big for the larger part. You could also make a goldberg-esq sorting slot, same mechanisms they use in vending machines to sort coins. Small hole followed by a large on a ramp. Small parts roll down and drop through first small hole, larger parts roll past first hole and into second large hole.
If the seals are too flimsy to permit these methods reliably then yeah, some sort of stamp or mark will suffice. I guess you could give the monkeys a piece of tube to measure against but that would slow the process down more than required I feel.
>>944090
how is this even possible?
you need to give us actual examples of this happening, show how you traced the cause of the problem and where it showed up in QC.
it could just be one shitstain worker fucking up your orders. it could be one guy in inventory not giving a fuck, it could be some guy thinking its funny to swap some from each bin.
when these are produced, they should be placed into seperate bins and stored or shipped out together. by the sounds of it, you're somehow getting individual packers to select products off a shelf.
you could do a stock take, count the seals, figure out who packed each job that fucked up and try to work out the problem. your first point of call should be segregating the imperial from the metric seals. they should not be anywhere near each other.
this is not a matter of oh well we should just make them different colors. that would increase production cost. you need to look at management and work out why this happened, what is wrong with your process control.
>>944090
The simple answer is different colors.
>>944105
Well, no, the simple answers are >>944100 followed by >>944098. Adding an entire new machine to add dye to the seal mixture (assuming thats even done in house) is a lot more costly than a simple ramp segregator in the short term, and does noting to stop the materials being mixed in the first place (which is what incurred the cost of the dying step in the first place)
Lrn2ProcessEfficiency anon.
mass firings. replace your monkeys with something reliable like chinese grandmas or a robot.
how can we trust your seals to meet spec if you don't even look in the fuckin box before you ship?
we can't trust our own monkeys to catch that shit before it destroys a customer's machinery, if it came from a packet saying 36 but it's a 37, the dumbasses will go to the 100 ton press and MAKE it fit.
From what I've heard, one of the primary issues has been mixing up seals that have the same OD, and similar but different ID's. So when the old mexican lady in shipping stacks them up and squints her eyes at them, they all look good.
>>944111
>implying I'm not a blackbelt and TQA certified
You spent most of your post talking about how it's likely a human error. You're also making the probably erroneous assumption that SAE and Metric aren't segregated in some manner.
This segregation is failing and part of how it's failing is employee discrimination. There are two possible solutions to that:
>eliminate humans from the equation
This carries with it a whole host of additional problems all of which carry the burden of process bottlenecking not to mention quite probably creating a fucking huge increase in the between cycle times by having another component that has to be set up for different product runs while having to retrain significant portions of your production staff to new maintenance requirements, new shutdown/startup procedures, new calibration techniques, etc.
All while *still* having the problem of it being difficult for both your employees and everyone in your product's lifecycle to tell apart the similarly sized items.
or
>make it easier for employees and everyone involved with said product to tell the similar items apart
Dying isn't the only solution, nor is there an "entirely new machine" required if it's mixed on site. If it's mixed on site there's already an aerator/mixer/agitator. You just need someone to dump an additional bucket of stuff into this on select runs, which would, at most require some safety measures be rethought. If you want to get fancy, a metered hopper isn't some mystical piece of magic equipment.
If you're not mixing onsite, then another color is retardedly simple.
The bonus addition to that is now everyone in the product's lifecycle can more easily tell these fuckers apart, from your employees to your sales people (if they're sold at like a Fastenal or somesuch), to your customers.
You may still want to remove the human element at some point, but you can only remove it from the factory floor. At some point you have to rely on people.
Normally, this should not be possible. Practically all seals made of elastomers are injection-molded. this means, that there is one tool for every size or type of seal. Injection-molding machines like pic related usually run only one tool at a time. This means, if all seals from one machine, and only those, are fed into a box, it is not possible to mix things up. Seals don't walk from one machine to the other and drop into that other box by themselves. The machine operator should be required to close that box once it is full, seal it so that nothing else can get in, and put his signature on it. Of course, the box must get a label with information about the exact type of seals. It must never be neccessary to determine the type of seals by measuring them, it must always be clear by the label on the package.
The solution to OPs problem should not be what to do in case of a mixup, it must be how to prevent mixups in the first place.
Also, start an employee idea system in your company. Usually the guys working at the very place where fuckups happen know best what to do to prevent them, but the management rather holds meetings among themselves instead of listening to them.
Already mentioned but use a color code them in some fashion. Even with a paint marker, I use them all the time.
Then the monkeys don't have to think twice but sort them by color, unless they're color blind. Then you're fucked.
>>944124
actually reading this, I wonder why management hasn't simply shit down the line onto the foreman.
>>944124
>Normally, this should not be possible.
This is a picker and packer problem I'm willing to bet that they run large batches of each size, then those get sent to somewhere where someone else has to pull orders. There could be some problems with the labeling of the containers/positioning of them (Bob likes his metric on the left, Tim likes it on the right) or maybe some manager tried to get creative by positioning bins by way of useage and stuck some similar ones together -- but long story short, at this point, and I can almost 100% promise it's at this point the fuckups are happening, people are getting confused over what's what.
>>944132
Either "progressive management" or it's been done a few times before and the problem persists.
>>944118
So? Close enough. Customers are too dumb to notice or care.
Pass\fail guage? May not work with rubber but I don't know.
>>944090
That is just fucked anon! You can't even get it right in-house anon? Fuck man. Label shit when it gets out from production! With a proper process that shouldn't be possible!
Your process is fucked if you get the wrong shit... It's not about labeling its about cleaning your process. You need to look for points of confusion and address those so that things like these aren't possible.
>>944090
>As for the issue...the production crew has recently gotten into the bad habit of mixing up seals. For exampl.e, one seal might have be 4 x 5 (ID x OD), and it'll get dumped in with a box of seals that are 100mm x 125mm. Very close but not the same. Now we're trying to find a fast and simple way for the monkeys in production to differentiate things.
do all the seals have different weights?
mebbe you could get sensitive scales that could automatically ID a seal based on the weight
>>944090
>.the production crew has recently gotten into the bad habit of mixing up seals.
>recently
>habit
Get rid of the crew member that fucks this shit up.
Check on CCTV what happens.
What is the dayly/per shift volume of production? How many monkeys?
Ok this isn't a diy thing, needs some sux smegma & lean applied! Issues like this start at the top and work down, management need to be aware it's an issue, foreman needs to know why it's an issue, front line need to know it's doing there job wrong
OP painful crap like this https://en.m.wikipedia.org/wiki/Toyota kata six sigma, lean & kaizen are things you'll need to understand to manage problems like this. Otherwise you'll fix the problem, "they" build a better idiot and you're back to square 1
>>944605
just dont mention the red bead game. management dont like to hear that the dick head down on the assembly line swapping parts willy nilly is actually their fault.
How are you losing control of the seals?
they go from molding machine to bin?
bin to shipping box?
is there some kind of off line sorting?
when you change over a machine, you should be clearing it 100% of all the previous parts run.
when you sort/pack it should be contained and sort one part number at a time. boxes sealed once full.
operator signature/part number/date on the box
all previous parts removed
you dont need markers/scales/vision or what ever
you need discipline at your plant and management that will follow through and enforce it
also six smegma and lean are just fucking buzz words and fads. they work great if all you want to do is a lot of extra paper work
>>944116
>chinese grandmas or a robot.
OMG how fucking true this is
Each type of seal should just be colored differently for organisation. I'm sure there are dyes that won't affect the manufacturing process for those seals or am I wrong?
>>944646
Maybe they ship a set of seals for a certain product and someone is picking metric/imperial when the other box is empty instead of going down and opening another box.
read up on Poka-yoka
a critical part of Toyota's production system, and a big part of the LEAN toolbox.
>literally japanese for "Foolproof"
>one of the reasons the yellow peril managed to get 100% quality on ALL deliveries as early as 1962
>>944090
Color code them as Inch or Metric
>>944653
Ah, idiots. It's always the idiots.
Go-no go gauge clearly marked with LARGE dimensional markings. Make from Delrin or other material that's hard-ish and light.
Delrin round stock is readily available and easy to machine and won't beat up seal IDs, but because seals flex your monkeys must be taught not to jam them. Gauge should have a flange so seal lays perpendicular to gage diameter surface.
Color coding is preferable, and you can also SEPARATE production runs of seals which may be confused with each other.
If they are marked, a magnifier can help sorting.
Track the problem to who is doing the wrong thing. You must have point of origin to fix your process by retraining.
>>944765
You can only go so far.
I work with a guy who let a trenching machine be seriously damaged because he could not be bothered to look at the decal near a grease fitting , reading "lubricate here til grease appears at flange, every ten hours of use".
Meanwhile he was claiming he's "lubed every fitting just like he should", as I showed the ruined bearings and carriers to the owner of the company.
Sometimes, all the manuals, decals, instruction and training videos do not work, because you have a shaved chimpanzee on staff.
>>944766
I don't like go-no-go gauges at these tolerances, the same monkeys that are picking the wrong ones to begin with won't care about the gauges. It'll just slow down everyone else that's actually productive. Might be a sneaky way to find them. Track worker productivity and see whos productivity doesn't go down after introducing the gauges, double check they're the ones fucking up, and deal with them.
some metric/imperial sizes you will not be able to control with a go no go
specifically 8mm and 5/16
and 16mm and 5/8
in regards to seals that will not be rigid enough
>>944090
Dye, for sure, 100% unambiguous, and dirt cheap
You have bigger problems though. Your production process should be set up so that this simply cannot happen. If your line ends in a big pile of mixed up seals then you have something very easy to fix.
Anon has already said it but lean management is something you may find useful here.
Point of failure could be picking orders from stock? In which case fucking segregate your seals baka, and dye them.
>>944129
>color blind
Then to fit in with the other manufacturing facilities you put the blind guy in charge of painting the seals initially
>>944090
Colour them during manufacture. It's an old process.
>>944119
>>implying I'm not a blackbelt and TQA certified
I had to take a class for that shit last semester. God damn do I hate you people. Fuck you all and your pointless circle jerk of jargon terminology and acronyms. I hope you all get trapped in an industrial press.
>>944090
It's already been said before, but can't be bothered to read the thread. A go-nogo gauge will slow everything down. Your best bet is to change colors, separate similar sizes physically in production(keeping the imperial as far as possible from metric), or to trace and stop hiring retards. Other solutions may be reasonable, but it's hard to tell without giving any more information that you'd want to provide about your man. environment.
so how the hell are seal A and seal B getting into the same area out of a container? they obviously aren't produced at the same line/area. why are individual pieces being removed from their process area?
>>945667
>I had to take a class for that shit last semester. God damn do I hate you people. Fuck you all and your pointless circle jerk of jargon terminology and acronyms. I hope you all get trapped in an industrial press.
Eh, I don't give much of a shit about it myself, I just know it's an easy as fuck job that pays well. What's kind of annoying about it is that the certs aren't transferable. Each company does its own, which also invariably (in my experience at least), means some retarded busness major has stuck his dick in it and completely undermined the data-driven nature of it and kills any real measurement ability the various programs have.
>>945678
It's most likely
>Order specifies X of whatever size
>order picker pulls a big handful, counts them into shipping box, returns them to wrong bin
Or something similar. I can't really see any other place, assuming the factory is set up at least somewhat sensibly, that it'd happen.
Of course, it also might not be that they're being mixed into bins. The bins might not be well labeled and Bob might be grabbing X thinking it's Y.
Third possibility is that lazy employee thinks there's no real difference and is just picking up whatever's close to the right size for an order thinking it won't be noticed.
>>945692
I don't want to derail too much, but I loved the statistics behind QA. I loved the analysis of efficiency pertaining to qualitative aspects of manufacturing.
But then there's a bunch of twats that have gone and made it a weird cult where everyone speaks in tounges, referring to the "5 S's" or the "some A.c.r.o.n.y.m." and "developed by TOYOTA!" with some tier system taken out of a dojo run by a failed art major.
That being said, it has been helpful in paying the bills while I finish college.
>>945762
>Be in college
>History major, but I'm a secret math nerd, taking calculus on the side at a community college because my Uni is such shit at math.
Anyway
>Nignog friend needs help with his business homework.
>can't figure out assembly line optimization problems at all.
>Business maths text book is retarded, near as I can remember it said to just use logic or random guesses.
>What the fuck
>This is a linear algebra problem
>Was only in linear algebra for a week
>Solve his entire homework assignment for the night in like 2 minutes
>Nignog gets As in all his classes because he copy/pastes from wikipedia.
