A Tale of Two Zippers

http://www.bunniestudios.com/blog/?p=4364

by zdw • 11 years ago

136 31 11 years ago

It's interesting to think of the perspective of wanting to make a product used by billions of people every day slightly worse so some labor could be made redundant.

eru• 11 years ago

Why? We do that all the time.

sukilot2• 11 years ago

What is the mirrored plastic he mentions at the end? Sounds like the horror we have come to know as "glossy" computer monitors.

Is it that awful fake chrome that looks like metal (and is called "chrome") until the consumer opens the package and realizes they've been conned?

derekp7• 11 years ago

I had an external Seagate hard drive that seemed to me that it had a high quality look -- it was a satin finish plastic, which resembled the plastic used on power tools. The next upgrade I got was a smooth shiny black plastic which I hated from the beginning, as it would get scratched up too much in my laptop bag. But guess which one most other people prefer buy?

jzwinck2• 11 years ago

Some Lenovo laptops have shiny black plastic lids, like the SL500 (designed over five years ago): http://www.notebookjournal.de/storage/show/image/image48aeba...

saurik• 11 years ago

Think of the plastic on an Apple wall transformer: that precisely flat and nearly reflective plastic finish.

Ellipsis753• 11 years ago

I find these kind of articles very interesting. It's interesting the play you get between "too expensive" and "worth the extra money".

It's interesting to see what's considered worth the extra effort.

aceperry• 11 years ago

Bunnie's blog serves as an important record of the tech manufacturing world today.

fcurella• 11 years ago

Reminded me of when I used to work at a zipper factory, doing exactly that job.

aaron695• 11 years ago

I’d like to imagine that most people, after watching a person join pullers to sliders for a couple minutes, will be quite alright to suffer the tiny bump on the tip of their zipper.

Would I like to sack someoone after seeing them at work..... Not so much. It's hard to sack people you have met.

Do I believe in efficiencies that will sack people I haven't met, sure.

Good article.

CephalopodMD• 11 years ago

why not use the holes in the tab to align/hang them from along some rod or track? That seems totally doable

joezydeco• 11 years ago

If it was "doable", it would have been done.

How exactly do you get all the tabs on the rod? Should the tab-forming machine put them on rods before transport to the press? Now you have to deal with rods full of tabs and keep the press fed, plus machine jams etc.

Look at how they're moving materials around the factory. A bin full of this, a bin full of that. Dump part A into hopper B.

Animats• 11 years ago

At 3:32 into this video

https://www.youtube.com/watch?v=cR4S8Bqjmsg

you can see the simple mechanism used on a similar vibratory feeder in a US plant to separate zipper pulls facing in the wrong direction.

There are common tricks used with vibratory feeders to separate parts. That part is wider at one end than at the other, and has a big hole and a small hole. There are standard vibratory feeder tricks to separate them.

Custom vibratory feeder accessories, though, have to be designed and built, which may be beyond this small shop. They're making very low end zipper pulls, made of pot metal (zinc and aluminum) and casting them. Pot metal is really easy to cast, but too brittle for that part. Expect those zipper pulls to break. (There are many videos on YouTube on how to replace broken zipper pulls.) Better zipper pulls are stamped from sheet metal - plated steel, aluminum, brass.

The video is from a small shop that's barely surviving. It sucks making small parts like that, because your supplier and customer are both bigger than you and can squeeze you on price.

YKK, the world's largest zipper manufacturer, ("Small part, big difference") offers the same pull, with no alignment projection. YKK leads in the business because their zippers are better. There are tough engineering problems in zipper manufacturing. They don't talk about those. There are lots of exterior pictures of YKK plants on the web. There are no interior pictures.

jzwinck2• 11 years ago

YKK's popular Vislon model has the tabs: http://www.ykkfastening.com/products/zipper/slider/zipper_ty... - yet other models from them do not (e.g. http://www.ykkfastening.com/products/zipper/slider/zipper_ty...). I wonder why.

abrowne• 11 years ago

The Vislon is a plastic zipper (i.e the zipper teeth, not the pull), so I'd guess cheaper than the other, which is for a metal zipper. Maybe that explains it?

HeyLaughingBoy• 11 years ago

Automated Material Handling is really an amazing subject. One of the machines I wrote code for had to get small cylindrical containers out of a bin, orient them properly, and then push them down into a receiving slot.

The amount of work the mechanical engineer designing the mechanisms went to, and the tiny, seemingly insignificant details that make such a huge difference in the end, were mindblowing. And this machine (a subassembly of a larger machine) had a target failure rate of no more than 1 jam per million cycles. Don't know that we ever achieved that, but we did come up with some clever automatic jam recovery algorithms. And then there was his description of the process the software should follow in order to properly dispense the units.

Debugging was interesting. I already had a fair amount of machine control experience before this task, but it was the first time I needed a high-speed video camera to debug my software! I discovered pretty quickly that in software I could optimize for throughput, or for reliability, but one was always at the expense of the other.

davidbanham• 11 years ago

This article isn't about zipper manufacturing. While the process of vibratory sorting is interesting, it's not the point Bunnie is making.

The point as I see it is that price signalling is a frustratingly blunt instrument.

It's annoying that so much of our species' output is channeled towards activities that improve precisely nobody's lives. This is an artifact of how hard it is to communicate throughout the entire stack in our modern, globalized, decentralized capitalist economies.

That's the bit to let your engineering brain get hung up on. Not vibratory sorter optimizations.

sukilot2• 11 years ago

I think you are overthinking it in this case. Things that look nice and feel nice sometimes cost more.

sls• 11 years ago

Did you read the article? He's not overthinking it, that's what the article is about.

icebraining• 11 years ago

I'm amazed reading your post, because that's not what I see here at all.

In fact, this shows how precise the signaling is. In any other economic system, you'd get one type of zippers (or even a single type of car, or house) because "that's good be enough for everyone."

But in this system, price signaling is so precise that it even allows for distinguishing between preferences in zipper designs, all the way from the buyer to the manufacturer on the other side of the world, who adjusts accordingly.

It's annoying that so much of our species' output is channeled towards activities that improve precisely nobody's lives.

But that's like, your opinion, man. The fact is that the people are spending more on smooth zippers, because they care about it. Who's to say their lives haven't been improved?

What you're arguing is that the system should optimize for the opinions of experts (such as engineers) rather than the opinions of the laypeople who actually buy the products. That has nothing to do with "bluntness".

shutupalready• 11 years ago

The tiny tab allows gravity to cause all the pullers to hang in the same direction as they fall into a rail toward the left.

The explanation above makes it sound like the pulls are getting re-oriented (flipped around) which is not what I see in the video.

I think a clearer explanation is this:

The pulls are aligned 50% tab-right (good) and 50% tab-left (bad) as they approach the feeder chain (or rail at the left as the OP calls it).

The tab-rights fit into slots in the feeder chain and proceed to the next assembly step.

The tab-lefts don't fit into the slots and simply drop away into the rotating pan below (from which they presumably return to the pot and make another attempt at the feeder later on).

yeahboats• 11 years ago

I think its more that the tiny tab prevents it from falling through the rail, acting as a pivot. Gravity pulls the rest of it down.

colanderman• 11 years ago

I like those little bumps. They make the zipper somewhat easier to pull (more friction = less grip needed).

lucio• 11 years ago

I’d like to imagine that most people, after watching a person join pullers to sliders for a couple minutes, will be quite alright to suffer the tiny bump on the tip of their zipper to save another human the fate of having to manually align pullers into sliders for 8 hours a day.

Leaving him in the street with no job...

The worker is happy that the designer don't like the "tabs". And yes, it is more expensive, because the required manual labor.

In any case, you'll be not making him "a favor" by automatizing his job.

I'm pro-automatization, but don't paint it as "a favor" for the ones which will be left without job...

nl2• 11 years ago

Leaving him[1] in the street with no job...

It's not that simple.

While I don't preclude that is a possible outcome, if you read the entire article you see that the fact that smooth zippers are more expensive means they can only be used on more expensive items, even though they don't give any practical increase in quality in terms of reliability.

If those smooth zippers were as cheap to produce as the ones with bumps, then it's possible (likely even? - after all, producers always want to differentiate their good in some way) the decreased cost of production would be invested in adding features that actually do something. One defensible possibility is "mass customization" type production, which relies on each piece being deliberately different in some way. I raise that possibility because most would consider that type of work to be "better" than manually threading zippers, and there could be more people employed.

Clearly there are a lot of things that could happen if both forms of zippers were as cheap to produce. My point is that it is difficult to say for sure the worker would be on the street with no job.

[1] It's more likely that the worker is actually female: Chinese factories generally prefer to hire women. It is only comparatively recently that the ratio pf production line workers has approached 50/50, and as recently as 2006 there were 14 men for every 100 women working on production lines in China. See http://www.ft.com/intl/cms/s/0/c8948f1e-2653-11e3-8ef6-00144...

zaroth2• 11 years ago

Obviously the manufacturer will try to optimize the BOM, and that includes the cost of the zipper. Just as someone spent an incredible number of hours assembling that plant, someone is also struggling over the BOM everywhere that zipper is used. And you can bet plants that make only indented zippers with even less labor exist.

What I came away with was first, the science of plant automation is far, far more advanced than computer science :-) Second, they've eliminated an incredible about of manual labor already.

Using factory automation as a model, it's interesting to think how software could do the same thing to the marketplace.

I agree with @nl, as long as the worker is there voluntarily, I don't see a problem with a market existing for sweating the small details. The upstream market competes by optimizing their BOM, so if the zipper isn't worth the cost, it will phase out.

To argue that it shouldn't be worth the cost because the job seems to boring for you is odd. There are many people who would be very happy to have such a job. If and when there is a future where there does not exist a large number of people would be overqualified for such a job, let me know.

timdierks• 11 years ago

Seems like a really good designer working with knowledge of manufacturing would design an aesthetically pleasing zipper pull with an asymmetry sufficient to allow such automated orientation (e.g. one end thicker across the width of the pull).

csandreasen• 11 years ago

Perhaps I'm missing something, but why not just put the tab on the other end where the wearer won't grab and pull? Perhaps there's a size/perspective thing going on with the pictures, but the tab looked small enough to me that it wouldn't interfere with rotating the pull around the joining mechanism.

leovonl• 11 years ago

I was surprised to find out this had nothing to do with Haskell.

A Tale of Two Zippers

Related Stories