I'm looking for help in finding improved methods for 99+% accuracy in bulk packaging 'small' assemblies (roughly 2 inches in diameter and 2.5 inches long). We have been tracking our errors for quite some time, but have not found our way to reducing errors. The major issue is incorrect quantity, with the secondary error of missing components in 'kits'. I would appreciate any reommendations or leads to websites, etc. :thanx: :bonk:

I'm looking for help in finding improved methods for 99+% accuracy in bulk packaging 'small' assemblies (roughly 2 inches in diameter and 2.5 inches long). We have been tracking our errors for quite some time, but have not found our way to reducing errors. The major issue is incorrect quantity, with the secondary error of missing components in 'kits'. I would appreciate any reommendations or leads to websites, etc. :thanx: :bonk:

I have seen "shadow boards" with a "vacuum-pac" work quite well. In fact almost idot proof. This application was used by a major manufacturer of x-ray and cat-scan equipment for their hospital installation kits. It was implemented shortly after they registrared to ISO 9001-1994. They photo printed the parts (including nuts and bolts) onto the cardboard backer. One photo for each part;i.e., if four bolts were needed, of a particluar size, then 4 photo's were put on the backer. These photo's were actually outlines of the parts. When fully implemented they never looked back problem was solved!

I am assuming you already use weigh counting for your bulk packaging. Is it possible your part is so light you might have to increase the "teaching" quantity to get a more accurate piece weight?

Dave

I'm looking for help in finding improved methods for 99+% accuracy in bulk packaging 'small' assemblies (roughly 2 inches in diameter and 2.5 inches long).


:read:

1] high resolution weight counter - cheap, sensitive to in door air flow ;

2] auto feeder + sensor/counter - expensive, tailor made for each parts , high precision (0 descrepancy til date in my experience).

I'm looking for help in finding improved methods for 99+% accuracy in bulk packaging 'small' assemblies (roughly 2 inches in diameter and 2.5 inches long). We have been tracking our errors for quite some time, but have not found our way to reducing errors. The major issue is incorrect quantity, with the secondary error of missing components in 'kits'. I would appreciate any reommendations or leads to websites, etc. :thanx: :bonk:
This is essentially an exercise in "mistake proofing." You might gain some insight into the process of mistake proofing by browsing this site:
http://www.campbell.berry.edu/faculty/jgrout/pokayoke.shtml
Jim Howe has described one form of mistake proofing. When you become familiar with the process of mistake proofing, you may be able to create one of your own.

Personally, I would not put much stock in trying to automate a process such as you describe unless you were doing more than 10,000 packages per shift. Automation considerations always come down to the economics of price per package when you compare

the cost of amortizing capital equipment and future changes to equipment to accommodate changes in product
AGAINST
the cost of providing a "mistake proof" process for humans.

We also had high shipping errors (low DPMs - but that is in another post) but high shipping errors. Quantity issues, wrong parts in the box etc. I used to receive approx. 3 - 5 complaints a day. One thing I implemented was a shipping inspection. Our lead inspector has to visually check all boxes going outthe back door. She is checking quantity and part content. After she has verified the contents she affixes a little sticker to the box. The sticker is bright green and states, "Inspected with Pride by Lupe." She also stamps the Sales Order as verified. This has reduced two problems. One - now our customers know we are checking every box and are more apt to "trust" us (issues from the past, do not ask!!) Two, we have traceability when there is an issue. We now have less than 1 shipping error per week.

The stickers have caught on as well. We now have "Packed with Pride by XXX"
"Manufactured with Pride by XXX" Actually now that I think about it our boxes of product look like a VW van from the 70's with all the bumper stickers (not really) :biglaugh:

We were in a high automation phase awhile back. The payback is just not there in our facility. In fact our old GM is no longer here, due to the fact that he purchased several hundred thousand dollars worth of automated equipment and received very little or no payback. They are nice and sexy to look at, however they are not achieving our goals, so it makes it harder for us to get the stuff we now need. For a good example of automation that worked there are several comapnies out there that have made it work after years of research, and process improvement. Nintendo in Seattle is one (ASQ conference ;) ) I recommend the tour if you are going.

Just some thoughts....

I'm looking for help in finding improved methods for 99+% accuracy in bulk packaging 'small' assemblies (roughly 2 inches in diameter and 2.5 inches long). We have been tracking our errors for quite some time, but have not found our way to reducing errors. The major issue is incorrect quantity, with the secondary error of missing components in 'kits'. I would appreciate any reommendations or leads to websites, etc. :thanx: :bonk:

Jim,

I know that you have stated that you have been "tracking" your errors, but to what degree have you performed root cause analysis (and be careful here that you're not looking for cause of each incident versus cause of the systemic problem)?
:bigwave:

We are having similar problems at our facility, and was wondering what people may have done to correct them.

Basically we have two frequent types of complaints, One is that we ship a container of parts that contains two different part numbers. Usually the bottom layer of the container has the wrong part, so even doing a top layer inspection would miss it.

The other complaint is just a container of parts that is mislabled. For one particular customer we make two different parts, both get packed in containers that are almost identical. basically what happens if that the containers are mislabled. This is something that a Top layer inspection would solve but my boss is looking for some sort of electronic system using scanners to track inventory and partials.

Right now we do have a computerized labeling system, but the labels are only put on finished containers. If a set up or change over is done before a container is finished, other than a "partial layer" tag there is no indication of what part is actually in the box.

Anyone have any Ideas?

We are having similar problems at our facility, and was wondering what people may have done to correct them.

Basically we have two frequent types of complaints, One is that we ship a container of parts that contains two different part numbers. Usually the bottom layer of the container has the wrong part, so even doing a top layer inspection would miss it.

The other complaint is just a container of parts that is mislabled. For one particular customer we make two different parts, both get packed in containers that are almost identical. basically what happens if that the containers are mislabled. This is something that a Top layer inspection would solve but my boss is looking for some sort of electronic system using scanners to track inventory and partials.

Right now we do have a computerized labeling system, but the labels are only put on finished containers. If a set up or change over is done before a container is finished, other than a "partial layer" tag there is no indication of what part is actually in the box.

Anyone have any Ideas?
This is almost a classic example of an opportunity for mistake proofing.

Possible solutions depend upon whether requirement to have two different parts in the same carton is a deal killer or whether you could have two separate smaller cartons (one for each part) then taped together and shipped simultaneously.

If the labels are computer-generated, let them be generated as part of the job traveler so they are available simultaneously with the parts arriving at final inspection and packaging to eliminate possible mixups.

Since you are relatively local to me, consider extending me an invitation to visit and look over your process for more precise tips on mistake proofing. This may even be an opportunity for a chapter in the book I'm working on:
Wes Bucey offers service as "Strategy Advisor" - Seeking Case Studies for Book
http://elsmar.com/Forums/showthread.php?t=9117

Thanks Wes,


I will pass your offer on to my Manager.

I was not very clear on the one type of mistake that we make.

nothing that we ship out of this plant should have more than one part number in a container.

What we think is happening is that a container may be partially packed, usually because of a product changeover or some sort of downtime.

The container is then improperly marked, and not labeled as a "partial"

and then is used again when another part number is set up, the new parts are then packed on top of the parts that were already in the container.

when the container is finally finished it is labled for the most recent parts put into the container, and then sealed. A top layer inspection would not catch this mistake, and unpacking to check the last layer is just not very feasable or desirable.

Basically what staff is trying to figure out is If there is a way to track a partial container electronically. So that they dont get "lost" or misused.

One of the reasons this happens is that we make multiple part numbers for the same customer, all the part numbers use the same returnable containers.

Right! One thing to consider is labels with line drawings of the part (goes along with preprinting labels to accompany part traveler) to avoid such confusion. That way, the line drawing and part number are prominent whenever there is an opportunity to put something in the container. If the part don't fit, don't do it!

Edit: added reminder to preprint labels.

Basically we have two frequent types of complaints, One is that we ship a container of parts that contains two different part numbers. Usually the bottom layer of the container has the wrong part, so even doing a top layer inspection would miss it.

The other complaint is just a container of parts that is mislabled. For one particular customer we make two different parts, both get packed in containers that are almost identical. basically what happens if that the containers are mislabled. This is something that a Top layer inspection would solve but my boss is looking for some sort of electronic system using scanners to track inventory and partials.
...
Anyone have any Ideas?

Recently, we've had to beef up our partial bin and empty bin inspections to ensure that they were being done properly. So many part numbers are similar and a cursory glance at a partial bin label might not catch errors. Now there is sign-off by a QA inspector and, in the case of partial bins, gauge checks to make sure the parts already in the bin and those being added are the same.

As for mis-labelled bins, we are going to on-demand printing of labels, tied to a KanBan system that ensures that items are finished in the correct priority. Label information may include colour code or other identifying information besides part number and destination plant. Final audit inspectors use PDAs that use the serial number on the shipping label to bring up part-specific inspection criteria and reference information.

B.G. Wiehle

Recently, we've had to beef up our partial bin and empty bin inspections to ensure that they were being done properly. So many part numbers are similar and a cursory glance at a partial bin label might not catch errors. Now there is sign-off by a QA inspector and, in the case of partial bins, gauge checks to make sure the parts already in the bin and those being added are the same.

As for mis-labelled bins, we are going to on-demand printing of labels, tied to a KanBan system that ensures that items are finished in the correct priority. Label information may include colour code or other identifying information besides part number and destination plant. Final audit inspectors use PDAs that use the serial number on the shipping label to bring up part-specific inspection criteria and reference information.

B.G. Wiehle

So by gauge checks, you mean that the inspector must take pieces out of the partial bin and inspec them to make sure they meet spec. That is an interesting Idea.

Wow your QA guys use PDA's out on the floor, thats pretty impressive.

Depending on the form, fit and function of similar parts, some organizations find ways to add non-functional markers (an extra groove?) to parts during manufacturing to make it easy to visually distinguish between parts without resorting to measuring with micrometers. We did this once on a family of parts identical except for 0.10 inches difference in depth on each of five parts in the family - the depth of a cavity which would accept a mating part. We were able to add grooves on the outside of the part which did not affect fit or function but allowed instant visual differentiation. We made similar grooves on the mating part and coincidentally eliminated mixups during assembly. Mating parts simply had to have the same groove set.

If it is inappropriate to change the form of a part, modern automatic marking machines can be added to a product line to individually mark each piece with a unique identifier. Pharmaceutical companies are doing it with barcodes on individual coated pills and capsules.

Depending on the form, fit and function of similar parts, some organizations find ways to add non-functional markers (an extra groove?) to parts during manufacturing to make it easy to visually distinguish between parts without resorting to measuring with micrometers. We did this once on a family of parts identical except for 0.10 inches difference in depth on each of five parts in the family - the depth of a cavity which would accept a mating part. We were able to add grooves on the outside of the part which did not affect fit or function but allowed instant visual differentiation. We made similar grooves on the mating part and coincidentally eliminated mixups during assembly. Mating parts simply had to have the same groove set.

If it is inappropriate to change the form of a part, modern automatic marking machines can be added to a product line to individually mark each piece with a unique identifier. Pharmaceutical companies are doing it with barcodes on individual coated pills and capsules.

We have tried to implement similar ideas as a way to mistrake proof other processes but Our customer (who shall remain nameless) has rejected any change regardless to how it affected fit or function. We always get "that design is locked and cannot be changed" from them.


I have been toying around with the idea of printing a unique bar code onto each piece.

We have tried to implement similar ideas as a way to mistrake proof other processes but Our customer (who shall remain nameless) has rejected any change regardless to how it affected fit or function. We always get "that design is locked and cannot be changed" from them.


I have been toying around with the idea of printing a unique bar code onto each piece.Of course, such anti-change reaction is common at every level of the supply chain.

As I have written elsewhere, I do not go into battle unarmed. Ultimately, you must be like the Godfather and make them an offer they can't refuse. On our groove situation, I wrote a detailed letter to chief engineers, chief financial officers, and chief executive officers of both my customer and his customer, explaining the added value, showing no detriment to the function, and offering to split my own savings with them. Who do you think wanted to go on record as declining the offer after seeing the list of other recipients at the bottom of my letter?

Some call it extortion, I called it success.

:topic: You've read some of my replies here in the Cove. Any question of how detailed (prints, physical samples, inspection reports) the letter was? Each recipient was forced to :ca: with each of the other recipients for fear of looking foolish if he didn't accept the proposal. I spent about $1,000 putting together the proposal and saved about $500,000 over five years as a result. That meant customers saved at least that much, probably more because they saved soft cost of mixups during assembly.. No one had to change the original drawings - merely issue an annually renewable waiver for the grooves on each part - costing customer less than $25/year for the entire family of parts. Work Instructions for assemblers were amended (revised) to add "alternate method of identifying parts when present . . ."