Hacker News new | past | comments | ask | show | jobs | submit login
3D printing and the future of retail (daltoncaldwell.com)
33 points by joeyespo on Feb 9, 2013 | hide | past | favorite | 19 comments



Interesting notions, but a few counterpoints:

+ 3-D printers aren't economical — Something like a drywall anchor takes a couple seconds to produce in a traditional injection molding process. It would likely take 15-30 minutes using an FDM 3-D printing process and even longer using a higher resolution SLA system.

+ Print-on-Demand hasn't taken off on books, why will it with physical goods? Print-on-Demand technology has been available for the better part of a decade, but Barnes & Noble and Borders never saw an opportunity to implement it. Again, this comes down to speed. For undifferentiated products, it will always be faster to mass produce.

+ Even taking the argument at face value that 3-D printing will help digitize inventory and create the same pressures that digital media did, I think more efficient mass production, rather than 3-D printing will be the likely case. Zara can produce limited runs of high fashion clothing and get them into malls all across the US in 2 weeks. Protomold makes low cost, small run injection molding accessible. We'll likely see more of this rather than fleets of MakerBots.

+ The problem with retail is the cost model of keeping store fronts open, not the way the products in it are produced. Even if we assumed there would be big box stores full of 3-D printers, they are still slow and do not have a Moore's Law like effect that will change that any time soon. If I ordered something at one of these 3-D printer stores, I'd have to wait hours to get my products. Why not just go straight to ecommerce at that point.

+ Producing Atoms is REALLY, REALLY hard. There are tens of thousands of different plastic formulations. Different additives, secondary processes, and assembly steps that go into high fit and finish products. Even the best 3-D printers can only handle a few materials — at most. There is no way to produce something like a pair of "Beats by Dre" headphones or a similarly complex electromechanical device using those systems. The simple things that can be produced would be more economical being mass produced and warehoused in a low cost location and shipped to customers rather than printed on demand.

Think about the difference in quality between a printer at home and one of the nice photosystems at a pro printer. Even then when dealing with paper stock and 4 colors of ink the difference is striking. Now increase the difficulty by several orders of magnitude.

3-D printers will have a huge impact, but more in the way open source programming languages did. They lower the bar to entry and let more people get involved with manufacturing, but will be limited utility tools, not everyday items — at least in the way this post posits.


Fitting additive manufacturing to the trajectory of books or CDs is mis-guided as the latter are ultimately vehicles for intangible goods. The recordable CD had an insignificant share next to commercially produced CDs when Napster and mp3 players made their debuts in 1999, shattering CDs and CD-Rs alike. Demand for physical and print-on-demand books is similarly crimped by digital books. Physical products have no equivalent - I can't sit on a CAD chair.


You're right in saying that 3D printing is a damn inefficient way to make a bazillion drywall anchors. But I only need one, and its a 10 minute drive each way in a 7000 pound diesel truck to go and get it.


Isn't it a little early to discount 3D printers? I remember in the late 70s, early 80s, dot matrix printers. Low-res, > 1 minute per page, monochrome, fanfold paper with feeds you have to tear off. Then it went laser, 300dpi, 600dpi, 1200dpi, 2400dpi, ink-jet, ink-jet color, prices went from $4k for laser printer to $80 for a scanner+fax+copy+color printer with auto-feed for scanning. Put photo paper in it and they'll print better than magazine quality color at 5760x1440dpi. Photo galleries now display ink-jet printed photos.

We're in that 1980s level of 3D printer. Give it 10-20 years and who knows. Maybe they'll be able to spew out 15 steel nails per minute or print 28 different materials. In fact given they can currently print organs maybe they'll be able to print wood. I have no idea where it will go but it certainly seems silly to think that today's 3D printers is the end of 3D printer tech. That's like my high school teachers who used to tell me printers would never replace typewriters because typing looked good and 9pixel low-res dot matrix printers looked like ass.


I'll counter some of your counter points :).

+ 3-D printers aren't economical — Something like a drywall anchor takes a couple seconds to produce in a traditional injection molding process. It would likely take 15-30 minutes using an FDM 3-D printing process and even longer using a higher resolution SLA system.

It depends what you're creating to determine how economical it is. A drywall anchor is obviously going to go down the mass production path as you need more than just one drywall anchor. A personalized iPhone case (with your name on it, to differentiate it from all your friend's iPhones) is a different story. It's significantly more expensive to mass produce truly personalized items ('truly' in the sense that the structure of what you're creating has changed - you could slap a sticker with your name on it, but that's a sticker on a case, not a personalized case).

I have an FDM printer, I'll print this (http://www.thingiverse.com/thing:5489) out and report back when it's done (it'll probably be about 15 minutes, but my printer is slow (40mm/s, some get up to 150mm/s). (before I was done writing this comment, my machine finished printing the drywall anchor - 10 minutes and 10 seconds of print time, see the picture here - http://imgur.com/a/iPnNx, it broke when I tried to use it (need to scale it up a little bit)).

The SLA - FDM argument isn't true, a higher resolution should take longer, but SLA systems can print 2D planes at a time (unless they raster, in which case their print head speed is limited by how quickly their mirrors can move), speeding up the process a LOT. I can set my FDM to a .01 layer height, but it would take 25 times as long to print something as opposed to using a .25 mm layer height. I can just sand the part in 5 minutes and print something new.

+ Print-on-Demand hasn't taken off on books, why will it with physical goods? Print-on-Demand technology has been available for the better part of a decade, but Barnes & Noble and Borders never saw an opportunity to implement it. Again, this comes down to speed. For undifferentiated products, it will always be faster to mass produce.

Your last sentence here is the reason 3D printing (if it does) will succeed. My sister broke the knob for her oven. It's an old oven, the company wanted $47 to replace the knob. They had a guy come over with a bunch of knobs, and for some reason none of them fit. I printed one, first try, and wham, it fit perfectly. No more pliers to turn the oven on! Now you wouldn't buy a printer specifically for this use case, but when my nieces and nephew come over and I can print them a glider airplane (that he took to show and tell) and a wand in the time it takes them to watch a show, that's pretty cool (they get a kick out of watching the part develop).

My sister brings me lots of stuff to fix (some I can, some I can't). I haven't done it yet - but having the ability to whip up obscure parts for 'disposable' toys allows them to stay alive a little longer. With a 3d printer, I can fix almost any mechanical problem you have with your toys.

It's not for everyone, but for the people who use them it gives them amazing abilities they did not have before. I don't have much money, so I can't get nice sheets of acrylic, wood, whatever to shape something out of, but I have a spool of plastic and can print almost anything.

+ Even taking the argument at face value that 3-D printing will help digitize inventory and create the same pressures that digital media did, I think more efficient mass production, rather than 3-D printing will be the likely case. Zara can produce limited runs of high fashion clothing and get them into malls all across the US in 2 weeks. Protomold makes low cost, small run injection molding accessible. We'll likely see more of this rather than fleets of MakerBots.

Thanks for the protomold link. I made a 3D printer for a product I want to make (hoping I'd be able to print some cases for it), but it's too slow for how many I want to sell. I'm debating 1/4" plywood, or injection molding. I think (as I have no money yet) wood will be the first run, but then I'd like to injection mold my parts (3D printing helps here too (or rapid prototyping), I can make sure my parts work before spending $1500 on a mold and crossing my fingers).

In terms of the actual article, I think 3D printing will compliment big box purchases, providing a market where add on customization (in many ways like a tattoo shop - you come in once, talk about what you want, set up a date when you want to get it done, and come back and the part is fixed / modified).

3D printing is what you use when you want something, but don't have a good way of getting it.


The main take-away I have from this blog post, is that the term "3D printing" is increasingly used to mean "automated manufacturing", even by people who are sophisticated enough to publish on Svbtle.

There aren't many products out there that can be produced by 3D printing alone. You can, however, get to the point the author suggests by applying "high automation" (integrating multiple automated fabrication methods into an automated manufacturing system).

Nevertheless, can at least those of us in the know insist on using the term "3D printing" precisely, i.e. only for additive manufacturing processes? That would be splendid, I think.


I'm with replicatorblog below. The article is sensationalist. And psadri is correct in that "on-demand manufacturing" is the real revolution.

From the article:

Q: "What percentage of the square footage of Home Depot is devoted to items that could cheaply and easily be 3D printed?"

A: Less than 1%

Here's some thoughts on the real issues in the space: http://www.nickpinkston.com/2013/01/some-thoughts-on-digital...


I spend hundreds of dollars each year at home depot on products that are good enough for what I need, but rarely perfect. I make things work, but they almost never have exactly what I want.

Hopefully there are enough people like me to make some kind of retail manufacturing on demand worth it.

Lowes has already switched to automated manufacturing for key duplication.


Oven knobs, fine. But there are so many things that can't be machine printed; they require the assembly of components made of very different materials.

Take "sporting goods" mentioned in the article. Tennis racquet: supposing you could print a frame that was strong enough (dubious) with an acceptable grip feel (very dubious), you can't print the strings. They need to be assembled into the frame. Baseball, baseball mitt, basketball, etc.

Somebody here mentioned you can't print headphones, anything electromechanical. I'll add, anything with a circuit board in it, anything with an LCD or OLED screen on it; anything with a motor in it; anything with a heating element (coffee maker, hot pad). Anything with springs or hinged or rotating parts (many toys).


Retailers aren't just financially leveraged, i.e. by debt, but also operationally leveraged, too.

A "business that has a higher proportion of fixed costs and a lower proportion of variable costs is said to have used more operating leverage" [1]. Given the low sales/fixed costs per square inch most brick-and-mortar retailers incur they would be vulnerable without a cent of debt.

Note that the prevalence of home inkjet printers hasn't bankrupted professional printers, just forced them to specialise.

[1] http://www.investopedia.com/terms/o/operatingleverage.asp#ax...


Retail won't go away. Shopping is something people like to do to some extent. Also, as someone mentioned, Starbucks thrives despite you could make your coffee at home. Even people who can cook well, go to restaurants... etc.


I think that there is a market for trying out products, but brick & mortar stores lose out on selection & price. I was toying with an idea for a business where there is a chain of locations that order in whatever you like to sample / try for a flat subscription fee. If you don't like it - it gets sent back or shipped to another location. If you like the product - you buy it for whatever is the cheapest online price. And, say, just walking into the store (with left-over products that weren't sent back yet) would be free.

This way, you get the best of both worlds: selection, price, and trying things out.


An interesting thought: coffee is probably much easier to make than arbitrary 3D objects made of various materials. Yet coffee makers have not become as ubiquitous as CD burners.

What I want to know is why.


Service, cleaning and maintenance maybe... the coffee from a cafe is prepared by a barista on a professional machine, perhaps tastes a bit better than the one at home, you don't have to clean the cups or equipment later and you don't have anything to maintain.


That's what I meant to say, why haven't the prices of those professional machines plummeted down to under $100? They're still very expensive.

You can imagine there being a machine that does all the cleaning, maintenance, etc. with a push of a button. It's just not as affordable as machines that burn 650 MB of audio data onto optical disks.


The cost of copper is the reason they haven't dropped in price. Those devices need to very precisely control temperature at high pressures. They are also designed to be able to make many expressos in a row while simultaneously outputting steam for frothing milk. This requires a lot of copper not only for the assembly that the water goes through but for a high pressure reservoir, which also needs to be made from copper.

The cost of the copper alone in some of those coffee makers is easily a couple hundred dollars. You also can't mix in another metal with a high specific head like aluminum in places to offset the cost because the presence of both copper and aluminum in the same water path will cause the two to react with one another.


I can make coffee with a £5 cafétiere that far outstrips what you get in a cheap coffee maker. Also, pretty much every PC with an optical drive comes with a CD/DVD burner these days. Coffee makers aren't attached to anything else.


see also that on-demand book printer that prints double-sided , cuts, and binds the book with cover. its on-site automated assembly in general. IKEA's strategy is a bit like this too. its like what if IKEA had a robot that manufactured the furniture at retail location? what if the lego block catalog was available from a kiosk in the toy aisle and you could print individual parts by tapping?


The internet is already doing brick and mortar stores in.

Perhaps on-demand manufacturing (for example, but not necessarily, 3D printing) will also contribute.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: