Back to Ganoksin | FAQ | Contact

Scanner - carving driver SW


#1

I am going to put forward a matter which I think is of general
interest, and ask those bcc’d to second it, so it will get posted.

About a year ago we watched different You Tubes as robotic machinery
scanned and then carved busts in stone, eg
http://www.ganoksin.com/gnkurl/ep7zof

Issue 1: Artistry)

Am I wrong that there would be a huge global market for this kind of
machinery if it the scanner -> finished product path were as easy as
going from the standard PC paper scanner -> finished product (paper
copy)e Some stone copies could be enlargements of originals and some
could be reduced in size as is done for paper copies. The technology
also allows us to go from one medium to another, eg wood -> stone or
vice-versa or clay/plasticine model -> stone. For example I do
amateur hand-building with clay and I can make a model of let us say
a fish in an hour. Using wood will take several times that and stone
carving it for me might take forever. I doubt that drawing the fish
on a computer screen with a CAD program would be as good as seeing
and feeling the 3D model.

Issue 2: Technology)

Robotic stone carving is dramatic, cutting edge technology. I was
amazed when I saw it first on You Tube especially with the
that the acuity of the scanning and carving both surpass
human visual-motor acuity. But experts have now advised caution in
purchasing this machinery because there is a “missing link”. That
missing link is perfectly analagous to being told that the paper
scanner for your PC has been invented and the paper copier to go with
it has also been invented but you do not know there is a missing link
which is the driver software. So you happily and naively buy scanner
and copier and then unhappily learn that you have opened an expense
account to make the two machines compatible which could bleed money
intil hell freezes over.

I would like to hear more on both of these issues. I will add only
this for now. Issue 1 does not detract from the artistry involved one
bit. Quite the contrary - the technology enhances the artistry in a
number of ways. Clay for example reaches a stage as it is drying when
it can be carved by wood-carving tools (my set has 24 picks and
blades) even though it has the look and feel of stone - in fact it is
stone, a very soft stone. But it allows you to make a lot of mistakes
without great expense which a piece of marble, jade or granite etc.
does not allow. Issue 2 is an investment issue. Most people on Orchid
are artists as well as investors. We invest in art with time and
money whether it is our own or someone else’s. IMO Issue 2 will be
solved soon if for no other reason than the fact that technology is
rapidly advancing on all fronts. Does anyone here have ideas on
where/when the driver SW will be developede


#2

Well, Peter, if frogs had wings they wouldn’t flop their arses when
they landed, you know? I notice there’s no reply to this, today, but
I’m just sitting here eating left over pizza… I’m far from a CNC
expert, but I know this much. CNC of all kinds needs to "know"
various things. The machine needs to know where it is (precisely),
where the work is, precisely, and what the tooling is. That part goes
for robotics like pick-and-place, too. The tooling isn’t just “a
forcep tip” it’s 28mm long with 6.85mm tips on 12mm tingers and a
throw of 8mm. And 180 deg rotation, all of which is programmed. And
the robot is precisely placed and the work comes at it, one way or
another, with precision so each part is always in precisely the same
spot. When you get into cutting and grinding, then the machine must
also know what the cutting tools ARE. Modern metal machining wouldn’t
be possible without indexable tooling (look it up…). And it also
needs to know what the material is and that doesn’t mean “metal”,
“Stone” or “plastic”. The program cycle for 1040 mild steel is an
entirely different thing from any stainless alloy which is entirely
different from brass. And you will note that the marble (butter,
relatively) in your video is mounted in a vise on a lathe bed. So.
You need to get a robot, which the easy part. Then you need to tell
the robot where it is, precisely, relative to the work, and fix it in
place rigidly. If you need to move it then you have to do that all
over again. Then you need to figure out how to devise a routine for
jade as a material, which I doubt exists right now. Then you need to
program all of your tooling, most of which will probably need to be
custom made, about how to cut relative to the properties of jade. You
see, you can’t just tell it “push”, it needs to have a feed rate at
all times or all of your efforts are wasted.

All of which will need to be custom made because you are the only
person on Earth that I’ve ever heard of wanting such a thing.
Meaning that you’ll never recoup costs from selling it elsewhere,
it’s just a custom program, for you. Then there’s debugging…
Now, I’m the one who deals in solutions, but this one is just one
big problem. If you had been out there with a grinder these last
couple of years you might actually have had some work done by
now…


#3

I received an off-list reply that this technology is not so “cutting
edge” (no pun intended). It is correct that you can Google on 3D
scanner-printer machines and they are widely sold and used (though
not as widespread as 2D scanner-printers). I would have no trouble
buying one and trusting the purchase for this reason. I can use
"references" as we do in personnel decisions, pick up the phone and
ask J Doe if the unit used in the J Doe office and purchased from ABC
company does the job

Differences in the receptor (probe/scanner) and effector (end
result) which may seem small to us are huge when it comes to a
technological solution. By analogy, object recognition is so mundane
to us that we rarely think about it. We “scan” an object with the
naked eye and say, That is a lamp and that is a chair and so on - big
deal. But it is a huge deal for machine developers and object
recognition is still far from a general solution.

The Next Engine scanner was also cited by another Orchidian
off-list. Have a look at the video which shows the machine scanning
the face of a live human (not a carved bust). If I got the WWW
correct, it is

Otherwise, click a few links on the Next Engine web site. That
interests me especially because the most popular local First Nation
carving for personal collections and investment (as they can increase
greatly in value over time) is the ceremonial mask.

One limitation of cedar however, is that it often cracks over time,
as happened to a valuable totem I examined a couple of days ago.
Stone masks would be more durable. But the artists are not going to
switch to carving in stone with hand tools and they are also not
going to become CAD artists.

Here we have 13,000 people and many must be interested in being able
to carve in stone with a robotic machine. The Carso lathe video says
the stone bust took only 8.5 hours including <1 1 hour of scan time.
How many Orchidians have scanned and carved an object of ANY size in
stone? I would very much like to hear from even one person who has
taken any object (large or small) and turned it into a stone replica
with an automated machine. It goes back to the matter of "references"
above. The robotic machine is like hiring a new human employee. If I
have no references who have worked with this person, I do not hire.
Do you?

The receptor part (scanning) seems well worked out. The effector
part is another matter. A 3D printer which builds up a model from wax
or plastic or whatever material it uses is not the same as a 3D stone
carving end-effector. A 3D end-effector which carves/grinds in wood
or plastic is not the same as a 3D unit which carves/grinds in stone.
Even then, consider the difference between carving in a marble with
hardness 3 and a calcedony river stone with hardness 7. Please do not
think I am being hair-splitting or an argumentative “troll”. Whether
we see these issues and deal with them or not spells the difference
between personal and financial success vs failure.


#4
Well, Peter, if frogs had wings they wouldn't flop their arses
when they landed, you know? 

Thanks for the inspirational imaginery JJ. Perhaps your baroque
flying frog will be my first figurine and no doubt a best-seller in
all gift stores.

The technical advice you gave corresponds more or less to the stages
we see in the video of the Carso lathe carving a bust over 8.5 hours.
They describe four stages. Each stage involves a hand-on human
intervention. As long as the seller explains the stages required for
the machine offered with a ball-park estimate of the time/cost for
each human intervention the problem is solved as far as management of
costs is concerned.

CNC means Computer Numerical Control… But the other NC in the
transaction is a cost accounting NC (also in binary though the
language is Dr/Cr rather than 0/1). Two others from Orchid have also
offered helpful advice off-list and it is up to them whether they
wish to join in the public discussion. That “human intervention” in
the stages of automation was also referred to. It may vary somewhat
from machine to machine.

It was also stated that this technology is not so new or cutting
edge. True as a generality on 3D scanning and printing. But not when
it comes to cutting/grinding stones with CNC machines. Have you, JJ,
ever completed a flying frog or other figure in stone with a CNC
machine? Has anybody else on Orchid? Nobody has come forward and
off-list I have only heard of two people who do this work regularly
and they are not in North America and there are associated
communication problems.

That takes me back to the analogy with a personnel decision. For all
important personnel decisions we get references and even if it is a
first-job situation the educational record of the new graduate is
the reference. However, I am close to making a purchase and mainly it
is that personnel reference that I await. If pick up the phone and J
Doe says, Yes I use the JJ CNC machine and it works very well with
stone, and better yet Doe says I will make a sample flying frog from
your material, then I am sure beyong reasonable doubt that I am
making a sound purchase. That is why I keep coming back to Orchid
with this topic. I am trying to “hire a robot”.

In all this time not one person out of 13,000 on Orchid has said, I
use the CNC machine regularly in my business with stone and I will
tell you about some of my experiences. Thus I conclude that this is
new and that those videos we see are mainly sellers’ demos.

One final point on your reply re “custom making” as my goal. That is
not the case. The original artists do the custom making in wood or
clay. I cannot expect them to do it as CAD on screen because they are
not that kind of artist. I want to then scan and replicate
(preferably with a software option to enlarge or make smaller) by CNC
machine. There are reasons other than commercial reasons for this.
Remember the story of Quianlong’s giant jade bowl? He had it hand
replicated which also gives some protection against theft or breakage
of the original.


#5
The technical advice you gave corresponds more or less to the
stages we see in the video of the Carso lathe carving a bust over
8.5 hours. They describe four stages. If we had some peanut butter
we could make some peanut butter and jelly sandwiches if we had
some jelly. 

Peter, you’re dreaming - I think most here understand that, which is
why only Andrew Werby, a real expert, bothers to reply here. I think
it’s useful info for many, is why I post. First some links:

http://www.ganoksin.com/gnkurl/ep7zpb
http://www.ganoksin.com/gnkurl/ep7zpc

Nephrite
http://www.ganoksin.com/gnkurl/ep7zpd

Also nephrite http://www.ganoksin.com/gnkurl/18w

The fluted top is nephrite http://www.ganoksin.com/gnkurl/ep7zpe

All of those were carved by me, and there’s some other pieces in
there I didn’t link to. I didn’t use a computer because ~no such
thing exists~. No such thing exists because it’s not necessary and
would be prohibitively expensive if somebody wanted it to make jade
trinkets. And I purposely pasted your quote about the Carso lathe
because you want to cling to that as though it actually means
something.

I’ll say it for the tenth time - marble is not jade, jade is not
marble. Marble is ridiculously easy to cut, which is why people have
done it over the centuries. Sculptors use steel chisels when
appropriate.

You actually CAN carve marble with steel or carbide rotary tools,
it’s just not appropriate. But it can be done. With nephrite it’s
just flat out impossible, which means that your lathe analogy is
utterly meaningless. Stop using it, it’s a dream. I’ll add two more
bricks to the pile, too. Robots are used for many things, but the two
most common uses I am aware of are pick-and-place and robotic
welding, which is big in the auto industry. Those robots are not
robust enough to put machining tools on the tips by a factor of
probably 1000. Put a real load on them and they will just cave and
flex. The other brick is that one pass of the carving tool is going
to remove .0001 (a ten thousandth) of material. That is fixed, it
can’t be changed for a given abrasive. So your robust robot also has
to be high-precision. And that high precision, robust robot needs to
mounted on uneven gound in the field and we haven’t even talked
about a power source. My guess, just a guess, is that you’re looking
at a million dollars and I’m not evensure it IS possible in any real
way. Sorry…