What doomed MakerBot? The Osborne effect
hugs.io
hugs.io
The frontier in 3D printing seems to be to get the cost down on the high-end processes, which are now very good. The Form1, the low-end stereolithography printer, is a nice little machine, and it's real. TechShop SF and Hacker Dojo in Mountain View both have one. Form1 charges $145/liter for the working fluid, so that's how they make their money.
Eventually, somebody is going to get low-cost laser sintering of metal powder figured out. But not quite yet. The Aurora Labs 3D metal printer turned out to cost about 10x the original $4000 price. MatterFab hasn't shipped their device. There's a low-end electroplating printer, but that's inherently a very slow process.
[1] http://gizmodo.com/why-3d-printing-is-overhyped-i-should-kno...
I think MakerBots biggest problem is that they've marketed (and with Gen5, designed) their products as being "consumer ready" when they're far from it.
With the huge brand name of MakerBot, I can't help feel that this is going to create a setback for the 3D printer market. It's just sad.
High quality desktop CNC Mills are out.
$3000 for a proven technology. Home-made CNC mills have always come in between $1000 to $3000, and now professional turnkey solutions are available in that price range.
I do realize CNC machines are the "opposite" of a 3D Printer. But if a CNC machine can get a software stack as easy to use as the 3d printers, then I think they'd get a lot more use.
Or if your RPM was too high and you smell burning, you drop it down on the next run.
In the worst case, if you catastrophically fail at those numbers and break say an end-mill ($20 to $50), that's still not that big of a deal. A major mistake / catastrophe still is cheaper to deal with than regular maintenance of say... a Form 1 stereo lithography plastic that costs something like $150 per liter.
I'd bet you that MeshCAM (again, free with the Nomad, $200ish otherwise) will create more precise parts than your typical Makerbot stuff anyway. Waterfall + Pencil Finish is more than enough to get the majority of projects done. (Steriolithography from Form1 looks... very impressive though, but the running costs are absurd in comparison)
ArtCAM, and other professional toolsets, are probably much better suited for professional artists who need to keep track of the grain of wood and the cutting direction. But... you can ignore the details of wood-grain and still come up with something with far more detail and precision than anything Makerbot can ever hope to do.
Basically, you start to care about those minor details with CNC machines because you actually have a machine that has such accuracy that those details matter. Also, $5 wood blocks and $20 leather (Drag Knife on a CNC machine: https://www.youtube.com/watch?v=bMoRUZnvbXw) honestly feel more "artisan" than the $100+/liter plastic that comes out of a Form1.
I'd really like to see PocketNC deliver.
Milling software remains a problem. For 3-axis machines and simple work, there's the Cut2D/Vcarve/Aspire line, which is easy to use. Most work at TechShop uses that program. At the high end, there's Hypermill - $22,000 a seat, but you get your money's worth.[1] Look at the videos of it controlling a 5-axis machine, with long thin tools working deep through narrow openings, without screwing up. In the middle, there's SprutCam, from Russia, which is startup-grade software - clever, but buggy. (I've used it.) Since in the CNC business, bad software results in rejected parts and tool and machine damage, buyers are unwilling to buy bad software.
There's some open source software for CAM, but it's below Cut2D, which is considered entry level.
in considering lowering costs i suppose the question is who is assessing these costs: an individual, or a company
form1 is interesting but at 3.3k that is quite expensive for an individual
there is an effort for a 100$ stereolithographic printer called the peachy printer
comparing specs:
pP(i) form1(ii)
resolution .2mm 20-200microns(.02-.2mm)
build vol ?inf?* 125×125×165
software open closed
price printer 100$** 3.3k$
price resin 60$ 150$
* peachy's design is formless, seemingly only limited by the tank volume and struts
** 100$ self assembly, 400$ fully built
peachy was impressive when it started its effort and has steadily improved but the really impressive element of the campaign for me is the transparency of development(iii), and the ethics of its staff(iv)
also the software is opensource : https://github.com/PeachyPrinter
.
(i) http://www.peachyprinter.com/#!printer-specs/c1gk6
(ii) http://formlabs.com/products/form-1-plus/
I don't speak for my employer at all, but I'm pretty excited that they should be out in the wild Really Soon(tm).
They have problems at two levels. At one level is that their management isn't very nice and thus the internal culture appears to be very strained. At least if you read the comments from about 50-60 former and current employees on one of those "employee satisfaction" websites whose name escapes me at the moment (Glass Door?).
They went from a very open company embracing open source and the Maker culture, to a closed company that was about pushing products that were not mature in order to meet revenue demands from their new owners. Their new owners being one of the big names in an industry that has been defined by patenting the fuck out of everything and then holding progress back for dear life. (High end 3D printers are the ultimate razor-blade business model -- with the slight difference that the razor handle isn't exactly cheap. Ask someone who owns a high end printer and they'll be able to tell you how insanely expensive they are to run).
At the more obvious level, Makerbot are in trouble because they have made a defective product. It would have been okay if the product was still "open" so users could service the parts or upgrade components, but now you can't.
In particular they have made an extruder that fails on almost every level. From the badly designed hotend that has an entirely wrong temperature gradient, to the non existing adjustment of idler wheel pressure, the unnecessary custom designed hobbed wheel, the lack of gearing on the stepper, meaning that it is too weak to do its job (look for "tick of death"), the fact that you can't open the extruder without voiding the warranty and risk breaking off the tabs to hold it together and the various Z-axis calibration problems which means that automatic offset measurement both results in the wrong distance AND it takes more time on EVERY print than it takes to calibrate the Z-offset on a manual printer ONCE.
The upshot of this is very frequent printing malfunctions, an extruder that is only guaranteed for 200 hours of operation before it has to be replaced, and if the extruder clogs, you have to send it in to have it cleared. (Extruder clogs can happen at any time. I've had extruders clog after just 2 hours). In Norway a new extruder costs about 4000,- NOK ($510)the last time I checked. For moderate use while prototyping you need between one and two new extruders per month, plus you need a population of extruders big enough to guarantee that you have extruders on hand when sending back extruders for repair.
You can calculate how many times over a year you pay the price of a new printer in spare parts alone. The Makerbot is a fucking expensive machine to run. (Yet not as expensive as the big high end printers from 3DS, Stratasys etc)
(Of course, in real life, time is worth more, so you fix the clogged extruders yourself and void the warranty).
So no, it is not about high end versus low end processes. Most people who are in the market for FFF (FDM) printers know precisely what the process can deliver. And there are printers that reliably deliver the expected results. Makerbot's fifth generation printers are not among them.
While the Osborne effect might have been a contributor to their earlier troubles, it isn't what is ailing them now. In fact, you would have to ignore half a dozen bigger issues before you would desperately reach for the Osborne effect.
(I have a Makerbot fifth gen at work and we've gone through three extruders before just giving up. The thing doesn't work. I occasionally update the firmware and the desktop software to see if things have improved, but there really isn't that much you can do when the key problem is horrendeously inept mechanical design.
I also have a Makergear M2 at home which (in Norway) costs about half of what the Makerbot costs, and which has worked impeccably for months now. After 4 monts the extruder mount started breaking apart. So I quickly printed a new one (before it broke completely) and replaced it. It cost me about $2 and it took 90 minutes from I discovered the cracks until I had the new part mounted. Apart from that I have had no problems and have had to replace no other parts)
It started when they initially refused to credit RepRap with all the open-source stuff they included in MakerBot. They then proceeded to bury the credit when they finally gave it.
They claimed they were putting a lot of unique IP in, but realistically in v1 it was just a different case.
Ultimately MakerBot tried to cash in on a trend, by taking a lot of open source design and building a product around it - but failing to do so in a open friendly fashion, and not understanding how to do product management.
But it is also really unfair to second guess someone who is riding a bucking bull in a nursery. That is what running a very successful startup can be like, lots of money coming in, decisions needing to be made with too little information, large outcome swings based on them with precious little runway to correct for errors. When you are in that space and someone reaches in and offers to lift you out, it can be hard to be rational about the choices you are making. It seemed to me that Stratasys knew exactly what they where doing, and less so for Bre and crew.
Unfortunately for Stratasys they failed to understand why Makerbot had been growing like it had, probably taking the late product launch as all the rationale they needed for a cash short company. I believe that had they understood what they had stepped in to, they would have approached it very differently and made very different choices. Not the least of which would be a mixed model of open and proprietary gizmos for their printers.
I am convinced that their lack of openness lead directly to the C.F. that is their "SmartStruder" which now has people calling for class action lawsuits.
A company with the Goodwill of a Printrbot and the resources of a Stratasys would have dominated the 3D market completely. Instead small printer companies are flourishing and kickstarters for new printers regularly cross $1M in pledges (and that is for printers no one has any right to believe can even be built!)
I found it hugely ironic that Bre is on the cover of this month's Popular Science [1]
Despite their looming fall and whether or not you feel betrayed by the switch to closed-source, IMO MakerBot played an important role in turning 3D printing from "a thing in movies and high-end research labs" to "a thing hobbyists and consumers could do". Their place in history may not be as large as they would have liked, but it's definitely there.
For example, I believe MakerBot changed their extruder design for one that was suspciously similar to something one of the Chinese manufacturers came up with. Then their parent company Stratasys sued that company to stop them selling their printers, claiming the extruder infringed a patent of theirs via some, errm, creative interpretation of the patent claims. Buying MakerBot printers doesn't even help fund the companies that do the original R&D - it helps fund someone that's trying to sue them out of business.
Their printers are not cheap, not reliable and not even all that original, so why would you buy them?
What is the story of this? (I tried searching for "SmartStruder class action" and all I found was this particular comment.)
The lack of real use-case is what's killing consumer-level 3d printers. Most people just don't need to run off a couple copies of a 3d trinket or toy with enough frequency to make it worth it. And the subset of those people with useful 3d-modeling skills is some tiny fraction of that number.
Whenever I see somebody using a 3d printer at the consumer level for something useful, it always seems to have just been for a one-off Arduino case or something like that. That's really just not incentive for me to spend the time and money to get setup with 3d printing, when I can just buy a case off of Amazon or whatever and be done with it.
The cost of the parts is the cost of your time + materials + setup costs (the printer). Is an Arduino case really worth $1400-$6500? That price only comes down if I print off more stuff I suppose, but at what point do the trinkets and dodads I'm printing off start to make economic and time sense? That's pretty far down the production chain, and I simply don't have that much stuff to print.
Also, your post reminds of a recent blog post by a friend of mind who is cofounder of a makerspace in the college town we were in. Loosely translated, he deplored how the maker culture has shifted from a utopia where everyone could learn to make their own durable and useful objects instead of buying into mass-production, planned obsolescence and consumerism, to a bunch of people manufacturing mostly useless plastic trinkets.
All in all though, when I move into a home with a garage, I'd love to get of those chinese mini-lathes which can be had for sub-1000$ :) And I wish laser cutters were less expensive (the trotec we had was bought around 40k I think, actually more than our CNCs).
(i) an arm on my glasses broke, i went to the shop i bought the glasses from, for 400$, and their suggestion? buy a new pair and my prescription has expired since the last time so i will need to have an eye exam for another 200$
(ii) the little plastic latch on my girlfriend's bike helmet broke in half, it still latches but can pop free with enough pressure
(iii) the clip on my compass broke so now i just tape a pencil onto it
(iv) variably sized plant pots
(v) mechanism holding the wheel on our closet door broke
(vi) clip on accessories: front and rear lights, phone nav; for bike
it's little annoyances that crop up and cause me to think i wish i could just print this little solution out
also..
(vii) i think parents would love the long term savings :: http://www.penny-arcade.com/news/post/2014/12/29/arts-and-cr...
(!) tinkering
.. 'all this and more' for 100$ .. http://www.peachyprinter.com/#!blog/c16fp
i think we need the printers in homes before we start to think about what we can print with them
the biggest problem i see is print time and recyclability of old parts
it is funny thought though.. i hate personal printers, all this useless paper lying about: "print me a reciept", "print me a boarding pass", "print me a copy of the report"
I am the founder of OpenPnP, an Open Source SMT pick and place platform. I eschewed 3D printing for my design work for several years thinking that it was not accurate enough, not strong enough, etc. Having experience with CNC mills I kept comparing 3D printed parts to ones cut accurately from aluminum and found them lacking.
What I recently came to realize is that the power of 3D printing is in it's instant turnaround and the ability to quickly iterate. Now, I design my parts for 3D printing. They look and function differently than they would if I had cut them from aluminum and I have to take the limitations of 3D printing into account, but they work. And most importantly, I can test a new design simply by hitting "Print" and waiting an hour or two.
I agree that much of the consumer market is dominated by people who buy a 3D printer and then use it to print trinkets until they get bored, but there are also a lot of people out there who are engineering new devices and machines using the ability to quickly test new designs on a 3D printer.
I have trouble shaking the feeling that printing 3D parts with little dabs of plastic has a lot in common with putting ink on pages with a dot-matrix printer. While they were revolutionary for the time, you don't see many dot-matrix printers these days.
The thing that a lot of people don't realize about traditional machining operations is just how long they take to set up. I didn't, myself, until I bought a CNC mill. I thought it was just a matter of sticking a big hunk of metal in it, hitting Go and then coming back when it's done.
The first problem is that for a typical 3D axis milling machine you have to find a way to mount the material you intend to cut without getting in the way of the cutter, and if your part is at all complex this can be very tricky. Different materials and thicknesses all require different methods of mounting, and learning how to mount and setup a part is an education in itself.
The second problem is that a 3 axis mill can really only cut down. That means if you want a hole in the up/down Z axis, you are fine. If you want one in the X/Y axes you are out of luck. You have to finish your Z axis operations, unmount the part, turn it, find a new way to mount it, zero the machine to the new part configuration and then do more operations. If you have a part that has, for instance, mounting holes on several sides you are looking at a LOT of manual work to complete that part. And each time you remount the part you have to be able to tell the machine exactly how the part is mounted so that everything lines up.
Industrial machine shops get around this by using 5 or more axis mills. These are insanely expensive and not really available to the home / hobby engineer, although I am quite interested in what http://www.pocketnc.com/ is doing.
3D printers, on the other hand, are truly 3D. You can have features in almost any configuration and it's no more difficult to print than a simple cube. It is literally a matter of starting it up and coming back when it's done.
I don't think 3D printing is the be all end all of machining, especially FFF, but for the home or hobby engineer it's an incredibly powerful and easy to use technology right now.
I'm not really disagreeing with you: right now it doesn't feel like 3D printers are a typical consumer item.
But experience tells me that this intuition is probably wrong.
As for "who needs it right now" I usually ask people if they own a plunge saw. Most people don't, but a lot of people do. In particular carpenters and DIY'ers. 3D printers are still unfamiliar territory for most people, but that'll change rapidly as they become better and cheaper.
In the future, as they become cheaper and better I don't see it as impossible that manufacturers of physical goods might use 3D printing to distribute spare parts and accessories for whatever they sell.
Hardware stores sell all kinds of DIY equipment, it seems to me that this is where they probably ultimately need to be channeled through.
Something that is a bit sad right now is that the printers you get from traditional retail stores are mostly the bad ones.
When I saw Makerbot make it onto the shelves in hardware stores I was a bit sad because I knew that this will lead to disappointment. I have seen a few other "brand" 3D printers which are not very good (and which try to develop a razor blade business model) as well.
I would also agree that they are a bit over-hyped. Just like home computers in the early 80s. And that there will probably be a backlash when people conclude that the Makerbots and the Dremels and the whatnots are really terrible products.
That's not the real problem. I guess it was a combination of chinese knockoffs and stratasys proprietary instincts.
In any case, IMHO, the layoffs are a result of losing the open community, which is now being serviced by true open-source proponents like Aleph with their excellent Lulzbot printers.
People forget when Osborne was losing sales, the IBM PC and PC-DOS was gaining sales and CP/M systems like those made by Osborne had a lot of competition. Microsoft/IBM had a converter program that could convert CP/M-80 programs to DOS programs. http://blogs.msdn.com/b/oldnewthing/archive/2015/04/17/10608...
Also Kaypro had won over Osborne customers with their own CP/M machines that reached the market before the new Osborne models.
Commodore had the VIC-20 and later Commodore-64 that provided cheaper computers as did the 8 bit Atari line. Some people would rather buy a $399 Commodore or Atari computer and hook it up to their TV screen.
So announcing a new product too far ahead of time was only one factor in the Osborne effect.
Makerbot has Chinese competition, and they had technical issues as well. There are more factors here than just announcing a new product too far ahead of time. Instead of Osborne it is more like the Apple III http://en.wikipedia.org/wiki/Apple_III
[1]: http://download.lulzbot.com/TAZ/
[Edit] LulzBot is more awesome than I thought. They post the source for future versions that they're working on, too: http://devel.lulzbot.com/TAZ/README.html "'Kauri' - EVT version of the next printer after TAZ 5.0"
OTOH MS vapourware has scuttled competitors. So it depends on your market power. Osbourne was against IBM and a disruptive tide.
With MakerBot, another issue may be that the market is approaching saturation. As they're being eaten by cheaper competitors, they haven't been able to create a new market to grow into. I suppose the Replicator is an attempt to do that - how has that done?
Gee, thanks, Bre. Again, why do you think that's true?
How did you come to that conclusion?
What do you know that those of us still in the "fantasy world" don't know?
Aha, that's it!
MakerBot fell victim to the Osborne effect, where talking openly about future products hurt the sale of their current products.
Sales dropped, so they got scared and went closed.
That sounds like a way more honest and real answer than the non-existence of an open source steel bender.
If I could have replaced the extruder with an open source clone that functioned better, I would have.
3D Robotics just raised $50M in a series C. It will be interesting to see how their commitment to openness holds up.
Agreed though, the magic is in the total sum.
The answer is your brand must must mean something. Be the domain expert. Be the one people turn to with questions. When the clones arrive, you are already ahead. That's easier said than done, and execution is everything and so on and so on. But there's not a vast difference between the open hardware space and other manufacturing businesses.
Also how can you be the domain expert if you open source everything you know? The millions you spend on R&D are given away and allow others to arrive at the same knowledge for $0. How does that work out for a company from a competitive standpoint? You are basically shooting yourself in the foot in terms of staying ahead and being a domain expert.
Either beat China at the manufacturing game, or continually be side-swiped by Chinese manufacturers.
Alas, no amount of open source anything is going to assist with that challenge ..
And instead build more robots. Local robots. Robots that build other robots, and .. things.
It is an emerging ethos, the local DIY-industrialist; actually a catch-up of the spirit that made China such cheap prowess in the first place, but when it happens as a Western phenomenon - i.e. local all the things. - it will indeed mean less cheap plastic crap floating around the ecosphere.
At least one can hope. An energy revolution and 3d-printing/transposing tech need not always start with a Fedex delivery. ("Solve global warming: stop using Fedex!")
If 3DR drones are more expensive, it's not only because they're not made in China. Manufacturing in Mexico is cheap enough, and buys you lead time, which suits their extremely fast iteration. A good idea in this gold rush.
People will pay more for the brand, if it stands for great support and integration. China is traditionally poor at those. 3DR started with a community, and regular people diving into drones will appreciate this comfort level.
TL;DR: To fight cheaper, make it better.
i have no stake in it myself, but if you're gonna make a strong statement like "doomed" you should back it up. layoffs != doomed
Link: http://makezine.com/2015/04/17/makerbot-lays-off-100-employe...
But you have a point, they're not dead yet. They might turn the corner, hire again, and regain that $100 million in value someday.
I learnt _huge_ amount pretty much accidentally from that machine and the user manual that came with it. Nobody told me at 12 or 14 years old that kids weren't supposed to solder together a handful of parts to add a joystick to the parallel port and write Z80 assembler routines to speed up the inner loops of games written in basic.
Taking it one step further and saying that it's the Osborne effect which caused their troubles is even less valid. There are a ton of different possible causes for MakerBot's troubles. The only honest answer here is that we don't know why MakerBot is having trouble, although there are a number of very good guesses we could venture.
The OP seems to be frustrated with MakerBot going closed-source. I share that frustration, but let's call that what it is, a frustration with MakerBot going closed-source, rather than using MakerBot's failures as an excuse to gloat.
we are in an epic financial bubble right now that DWARFS all previous bubbles of the past 20 years so much so that janet yellen herself said "cash is not a good store of value"; meaning the bubble is so big that those who created and sustain it with unlimited cash are so scared of it popping that they will publicly allude to the possibility of unending limitless printing ( and the inevitable hyperinflationary boom it would conclude with ) .
all of the self congratulatory lying and self deception of silicon valley is that it is 'libertarian' and 'independent' , when in fact ALL THE NEW MONEY IN SILICON VALLEY COMES FROM THE FED THROUGH EITHER BANKS OR THE TREASURY BY WAY OF VC AND THE MILITARY RESPECTIVELY.
the boom will end as all booms do and the headline of what 'doomed' makerbot will seem like a joke.
laying off 20% isn't 'doom' by any stretch of the imagination.
if there are problems with makerbots business model, then they were relegated to the parent company which now owns it.
bri pettis is a political genius for ousting his 2 compatriots and keeping the control, and the spoils of the buyout , all for himself.