RE: HP getting into the rapid prototype printer business

From: Charles Overy (
Date: Sat Sep 20 2003 - 03:49:12 EEST

ooohhhh, I am sooo misunderstood.

It is not that I do not believe that it CAN happen. I have very little
doubt that the technology is at HP and other places to make a $1000 printer.
In addition, I am sure that there are materials out there that can make a
retail end user cost effective part that is aesthetically pleasing and even
possibly useful for limited practical applications.

However I do not think that the consumer RP WILL happen until some
fundamental issues and technologies are addressed. Most of these are only
tangentially associated with RP technology but fundamentally associated with
RP adoption.

IMHO, there are some fundamentals that need to be in place before we see an
order of magnitude decrease in both machine price and cost per build/part.

Your analogy, Larry points out one of the biggest areas.
Last time I checked EATING was sort of essential to living.
Therefore a great many people do it every day
this results in a very large total market volume.
A lot of these people will try a great many variations on the eating thing
both to cut costs, or increase satisfaction, add variability etc.

Again, last time I checked, outside of this list and a few other odd places,
drawing in 3d is not essential to living.
Neither is the creation of instant replacement stove knobs
Or the creation of personal 3D sculptures
Or bookends of your families heads
Or edible babies from prenatal sonograms
Or the creation of sample engineering parts
Or the creation of custom jewelry
Or the creation of medical models
Or the creation of architectural models

In fact ALL of the items above, even when combined, represent a miniscule
fraction of the total market volume of the foodservices industry. Also I
would wager, that combined they represent a very small fraction of the total
market volume for photographic prints, digital or otherwise.

In order for there to be widespread adoption of RP there has to be much
wider use of 3D computing. It would be interesting to look at things like
what percentage of 3D data currently gets output on any sort of RP machine,
ever. Also:
What percentage of 3D data is it reasonably feasible to output on an RP
What percentage of total design is done in 3d
Of all the volume of computers sold in say 2003, how many have any sort of
3d applications actually used on them (Including VRML, VET, etc.)

I do not know the answers to these questions but I suspect that they are all
fairly small percentages. Larger, and growing for web 3D like VET but still
pretty small. Therefore, even if there was a $1000 printer that could make
a $10 part, I think there would be very few people that COULD take advantage
of it even if they wanted to. If you look at the displays, discussions,
etc at SIGGRAPH, even where the parts are going for FREE, there is far more
pure virtual work and 2d art than there is 3d. (Although there has been a
significant increase in RP works displayed including some by list
participants like Bathsheba and people like Carlo Sequin.)

Frankly, I do not see RP as the "killer ap" that will drive 3D adoption.

There are other technologies that I see as critical parts of that equation:

3D data acquisition and processing (Robust NURBS creation using point clouds
is starting to come out of the stratosphere but is still very labor
intensive)> There was a very interesting paper this year at SIGGRAPH that
elaborated on the "consumer" 3D camera attachment that had been announced at
previous conferences. Basically they were able to use a LED scanning laser
module from Symbol Technologies, in place of a flash. They created a module
that fitted to a standard digital camera flash shoe. When the Camera was
fired, the "flash" first put up a red mesh on the subject and a picture was
taken of that, then the camera took another regular picture. The teams
software then created a 3D image from the distortion of the grid and
overlayed the second photo as a texture map. Post processing was fully
automatic and took only seconds. While the approach had some problems it
demonstrates the type of 3D data acquisition that I think is necessary for
consumers to want 3D output.

Interactive or intuitive 3D content creation. Many many people have tried
to crack this nut but I still think it is beyond the ability of the
mainstream consumer to create a useful part or a interesting part. Emphasis
here is on MAINSTREAM, i.e. enough people need to be able to do this to
create very significant market volumes. It is disappointing, but telling,
that there has been very little adoption of any input devices or viewing
devices that are not mainstream WIMP GUI. Blame it on Mircosoft, blame it on
whomever you want but what happened to Spaceball, Sony Glasstron, etc.
Weren't we all supposed to be working in fully virtual environments by now?
Haptics are seeing almost no consumer adoption outside of gameplay despite
conclusive evidence that even the most basic tactile feedback greatly
improves the efficiency of computing. Does microsoft even make its tactile
feedback mouse anymore?

64 Bit computing breaking the 4Gig RAM barrier. Voxel computing enabled by
the above allowing a new class of 3d "solid" content creation.

Anybody have any others?

There is significant progress on all of these fronts but no unifying market

Also, there is my long term stump that the RP community needs to get
together and burry, STL once and for all. Kill it dead, drive a stake
through it if you have to. How about minor steps forward like a file format
that does not allow the geometry to fail because of rounding errors. How
about every vertex only ONCE.


A happy weekend to all

Charles Overy

> -----Original Message-----
> From: []On
> Behalf Of Blasch, Larry
> Sent: Friday, September 19, 2003 12:51 PM
> To: '';
> Subject: RE: HP getting into the rapid prototype printer business
> Charles and list,
> None of the current RP machines are designed to produce components using
> processes with any less complexity and mess than the average meal made in
> the average kitchen in the average house.
> When you make dinner, you start with lots of ingredients
> (expensive or not)
> and pots and pans and bowls and spoons and you mix and cook and combine
> things to produce the ultimate short term consumable. Then you wash and
> clean and put things away only to do it again at the next meal.
> Granted, the materials are somewhat less nasty, If you ignore
> salmonella and
> botulism...
> People spend outrageous amounts of money buying convenience foods that are
> either fully prepared or partially prepared to save time and make
> their life
> easier. Often, the price that they pay for the convenience is
> several orders
> of magnitude higher than the cost of the un-prepared "raw"
> material but the
> convenience is worth the extra cost. (It must be or else they are just
> stupid with their money.)
> How does someone determine what constitutes a convenience, and produce a
> product that represents that to the consumer? You must evaluate what the
> customer considers to be a waste of time and design a product or process
> that can be marketed with that benefit in mind.
> Most of the gadgets sold today are just that. Just think; people actually
> pay extra for the convenience of holding a redundant switching system for
> their TV in the palm of their hand. I remember when they came
> out. (yup I'm
> that old) A TV set with a remote control cost $50-100 more than a standard
> set. Now they are standard and many of the control features are much more
> difficult to use on the console itself.
> All this leads to the real issue, What is the convenience that can be
> provided by a home RP unit?
> It's not a necessity... after all, the items produced can probably be made
> some other way for less money (in quantity) right?
> Sincerely,
> Larry Blasch
> Lawrence R. Blasch
> Design Engineer
> CAE Systems Administrator
> OPW Fueling Components
> P.O. Box 405003
> Cincinnati, OH 45240-5003 USA
> Voice: (513) 870-3356
> Fax: (513) 870-3275
> -----Original Message-----
> From: Charles Overy []
> Sent: Thursday, September 18, 2003 7:00 PM
> To:; Blasch, Larry;
> Subject: RE: HP getting into the rapid prototype printer business
> Sorry the end of that message was
> If you look at
> you can actually get some numbers on this idea of parts bureau down the
> street.
> Stove knobs range from $7 to $30, with the modal cost of $25.
> That is going to be tought to beat.
> Charles
> > -----Original Message-----
> > From: Charles Overy []
> > Sent: Thursday, September 18, 2003 4:57 PM
> > To: Blasch, Larry;
> > Subject: RE: HP getting into the rapid prototype printer business
> >
> >
> > Larry,
> >
> > I think you have hit upon a very critical issue, the cost of
> > materials. However, the problem with recylable materials is
> > that injet cartidge/RP supplies are a major component of the
> > business model for these companies. The revenue stream generated
> > by the reoccuring sales of suppies is, at least for HP, greater
> > than the gross revenue from product sales and the margins are
> > much better (lower shipping costs, lower R&D, lower cost of
> > selling, lower tech support, little or no software development).
> > That is NOT to say that everyone is getting screwed on their
> > supplies, basically you pay for the machine technology, and lower
> > cost of future generations as you go. I doubt very much that HP
> > would be selling $100 color printers at Walmart if they could not
> > get a substantial number of those customers to plop in a $35
> > cartidge every few months. It is certainly why HP (and others)
> > have continued to improve the printers so that they will do
> > photos. Photos use TONS of ink sold at the retail level, that is
> > why they throw in the software to help you print many copies of
> > your photos.
> >
> > Back to RP, at any given point in time we probably have $4000 in
> > RP supplies and materials on hand. For an SLA it could be a
> > whole lot more and I am guessing more as well for a
> > sinterstation. To make a model of any reasonable size and
> > interest costs us say at least $100 in consumables. I belive
> > then that you need the same order of magnitude reduction in
> > material costs as you do in machine cost. Even then it will be
> > pretty expensive for the home user.
> >
> > If you look at
> >
> > > -----Original Message-----
> > > From: []On
> > > Behalf Of Blasch, Larry
> > > Sent: Thursday, September 18, 2003 8:33 AM
> > > To:
> > > Subject: FW: HP getting into the rapid prototype printer business
> > >
> > >
> > > Sorry about this being a forward... I'm having trouble
> submitting to the
> > > list due to my own SPAM filtering software.
> > >
> > > -----Original Message-----
> > > From: Blasch, Larry
> > > Sent: Thursday, September 18, 2003 9:16 AM
> > > To: ''
> > > Subject: RE: HP getting into the rapid prototype printer business
> > >
> > >
> > > Dear RP-ML...
> > >
> > > It does not matter what the material or process actually is, that
> > > is used to
> > > create items for the homeowner, until the process is reversible or the
> > > material recyclable, it will never become mainstream.
> > >
> > > This is not an issue of "Green" politics or ecology, it's a matter of
> > > practicality.
> > >
> > > The average "Joe Six-pack" won't want to stockpile a quantity if raw
> > > material at any price, and wouldn't be happy about throwing away
> > > personally
> > > created parts.
> > >
> > > Think I'm wrong? OK, how many people keep a supply of inkjet
> cartridges
> > > around the house? Do you think it's too much to keep an extra
> > $30-50 worth
> > > of supplies handy? How is this potential home RP printer
> going to change
> > > your mind? If the supplies are dirt cheap relative to the current
> > > SLA/FDM/SLS/inkjet materials, people would still complain that it
> > > costs too
> > > much to keep supplies on hand.
> > >
> > > Besides, the actual raw material cost for many engineered items
> > > (You know...
> > > the stuff that people are supposed to be using the home RP
> > > machine to make.)
> > > is pretty high, especially in small quantities. Unless the
> > > homemade RP parts
> > > can be recycled, or reused, the machine will never be more than
> > a hobbyist
> > > toy.
> > >
> > > Now look what happens if you make the process recyclable... You
> > throw the
> > > broken part into the machine and it gets re-processed into a
> > new part. All
> > > that you need to add is energy and perhaps an additional supply
> > of generic
> > > raw material/binder. Now the process will be accepted at home
> > and the raw
> > > material issue is nearly eliminated.
> > >
> > > The potential for RP to replace the inventory of "obsolete"
> > > repair/replacement parts for any durable goods store is pretty
> > > good. Imagine
> > > your refrigerator door handle breaks, you call the store and
> > they build a
> > > replacement on their big, high speed replicator from the 3D
> engineering
> > > design data library that the manufacturer offers on line (for a
> > > subscription
> > > fee). You just pick it up or have it delivered. The raw
> material is now
> > > purchased in large quantities and the manufacturing cost is
> > offset by the
> > > elimination of spare parts inventory. This applies to almost any
> > > material or
> > > process so it could apply to many different industries.
> > >
> > > Most of the materials used in products today are produced in
> > > large quantity
> > > to supply an existing infrastructure that expects to make things by
> > > softening and re-shaping small units of the actual material
> > > without changing
> > > it's chemical structure. This works for a business model that
> uses high
> > > volume, production tooling, dedicated process equipment to
> produce large
> > > quantities of the same part or product. The term "Raw Material" is not
> > > really correct in this model, since you are just processing
> an un-formed
> > > material into it's final shape.
> > >
> > > The manufacturing processes used to produce the plastic molding
> > > pellets are
> > > not even starting with "raw materials" since they often work
> > with refined
> > > oils, not the crude out of the ground.
> > >
> > > I'll propose another approach to RP materials... How about a
> > > machine that is
> > > hooked to your natural gas line and converts the methane into
> plastic as
> > > needed by the RP process? How many items can you make from
> > > different grades
> > > of polyethylene? Raw material flows in through an existing supply
> > > system and
> > > you just pay the bill every month based on the amount used. Oil
> > > or gasoline
> > > would work also, but they're not as convenient.
> > >
> > > My humble opinion...
> > >
> > > Larry Blasch
> > >
> > > Lawrence R. Blasch
> > > Design Engineer
> > > CAE Systems Administrator
> > >
> > > OPW Fueling Components
> > > P.O. Box 405003
> > > Cincinnati, OH 45240-5003 USA
> > > Voice: (513) 870-3356
> > > Fax: (513) 870-3275
> > >
> > >
> > >

This archive was generated by hypermail 2.1.7 : Sat Jan 17 2004 - 15:18:08 EET