#163 from R&D
Innovator Volume 4, Number 6
Dream That’s Coming True
Uziel is founder, president and chief executive officer of Soligen
I love history,
and have always been fascinated by the influences of great inventors.
As a child, I dreamed of being in the same category as
Gutenberg, Edison, and Newton.
But it didn’t take long to realize that I wasn’t
exactly in their league. However,
I’ve never lost the desire to contribute a highly significant
Back in 1981, I
worked for Optrotech, Ltd., an Israeli company that pioneered the
automated optical inspection of printed circuit boards.
Eight years later, the company was the recognized
market leader. Then,
I was vice president of new business development, looking for new
simple idea of a printing press using moveable type resulted in the
printing revolution. It
made the written word available to everyone.
By analogy, I was looking for a way to create a 3-dimensional printing press using moveable type.
I envisioned a machine fed with
computer aided design (CAD) information for a manufactured
part—say an automotive engine block. The machine would automatically, and simply, produce the
At that time, I
didn’t have a clue of how to make such a 3-dimensional printing
press. When I shared this idea with my close friends and business
partners, they became concerned that I was having an early
for Parts Manufacturing
components (e.g. tools) and functional parts (e.g. pistons) are
either made by carving them out of billets or by casting them into
pre-made cavities or molds made of either steel or sand.
The steel molds, themselves, are made by casting steel into
ceramic molds. A
3-dimensional printing press that automatically produces ceramic
molds, could be useful for both direct metal parts
fabrication, as well as casting the steel molds, themselves.
process of making ceramic molds for metal casting today is only
slightly different from the process which the Romans used two
thousand years ago to make tools and vessels.
That process used a wood or a wax pattern (the
positive impression) of the item, which was covered with sand or
clay. After the clay
mold (the negative impression) hardened, the wood pattern was
pulled out of the clay mold or the wax was burned out.
Molten metal is poured inside to form the final object.
After the metal solidifies, the clay mold is cracked
open to yield the metal part. A separate clay mold was
therefore needed for each different metal part.
pattern-makers are specialists who form their patterns from
the wood has been replaced by other materials, pattern-makers
often use manual sculpting to achieve the desired shape and
texture. It’s a
slow and expensive process. While
the expense may be justified for an object manufactured in large
quantities, it really becomes costly for parts that are produced
in relatively small quantities.
The expense is especially felt for prototypes that must be tested--and
the commercial stage.
My first stab at
fabricating patterns automatically took place at Optrotech, where
I built an apparatus that could create 3-dimensional objects
directly from CAD files. Using
electro-optics, we controlled a laser beam to draw cross sections
of the pattern on the top layer of an optical liquid glue
that solidified under the laser.
We made a succession of cross sections until the entire
pattern was formed in
solid, laminated plastic.
That plastic pattern became the master, or positive
impression, from which ceramic molds (negative impressions) could
To test the
machine’s ability to make a complex part, I programmed the computer
to form a chain, which is excruciatingly difficult to produce with
standard mold-making techniques.
It worked beautifully, and I could hardly contain my
excitement--this system was powerful!
partners tried to persuade me to abandon the idea of making
patterns for metal casting, and focus on making plastic
prototypes. They said
that the market would love a 3- dimensional printer for design
verification, even if the plastic parts are just models and could
not be tested functionally.
They also pointed out that most foundries were poor, "mom
and pop" shops which use primitive technology, certainly not computers.
And they have no incentive to risk their little capital for this
I knew that
bringing the technology to market would take several years and
many millions of dollars. Still,
convinced there must be some interest in direct pattern making for
metal casting, I visited more than 50 foundries around the
world. My partners
were right: almost
all foundrymen were set in their ways.
Our rapid prototyping excitement didn’t last long either. While attempting to patent my new invention, we learned
that a California start-up company called 3D Systems was planning
to commercialize a similar idea.
3D Systems was
interested in rapid prototyping, and agreed to form a joint
venture with Optrotech to use our technology.
I was hoping that, after a while, I would be able to
resume my rapid pattern making as a separate avenue for rapid
prototyping. However, despite my relentless efforts, the project flunked
before Optrotech’s funding committee and I felt that I missed my
opportunity to become another Gutenberg.
In 1989, I joined
3D systems as vice president of engineering. After 4 months of concentrated effort, we successfully
launched the rapid prototyping system.
However, I still saw rapid prototyping as only an interim step. My vision
was to produce parts--directly.
At that time, a
team of engineers at the Massachusetts Institute of Technology
(MIT) led by Dr. Emanuel Sachs and Dr. Mike Cima, was developing a
process to create ceramic molds (the negative impression)
directly. Their idea
was to move an inkjet-type printing head across a fine layer of
ceramic powder. Under control of a computer, the head would “spit” tiny
drops of liquid binder which would solidify the powder to
make the cross-section. As
with the laser rapid prototyping system, repeated cross-sections would
be assembled to produce the product--which in this case was the ceramic
mold itself, instead of the pattern the machine could directly
produce a ceramic-mold with no geometrical limitation.
At 3D Systems, our rapid prototyping focused on the master
form (positive impression), but the MIT process completely
bypassed it, making the mold (negative impression) directly.
Here was an opportunity to revive my Gutenberg dream,
I left 3D Systems
to pursue this idea at my own start-up, Soligen, and bought
the license to apply the MIT concept to the metal-parts industry.
We planned to initially capture the “high end” part of
the market for metal parts (parts with complex geometry made in
small quantities). I
recalled the cautions of my ex-partners in Optrotech, but I just
knew we could do it—if we could only fund the development.
We finally persuaded several companies to advance money so
we could build our first machines.
We called the new concept Direct Shell Production Casting (DSPC).
In June of 1992,
a handful of engineers and I worked around the clock in my garage
to design, build, and install three Alpha machines. By October, we demonstrated our concept (without even a fully operational
Alpha machine) at the annual Investment Casting Institute.
Our demonstration attracted attention and we raised more
Before we moved
to Soligen's current site, we had two interesting visitors.
One was Bill Torodoff, head of the casting division of
Ashland Chemical, who read about Soligen’s DSPC technology and
couldn’t believe that it would work.
Shortly thereafter, Ashland became an additional
co-development partner and this gained us substantial exposure.
The other visitor was a detective--summoned by a suspicious neighbor
who wondered about the strange activities at my house.
The detective, happily, was an ex-engineer, and he let
us complete our work--and visited regularly to see our
progress--even though we were breaking a zoning ordinance.
At that point we
had to decide on our marketing strategy, should we focus on
developing the DSPC technology and sell the machines to foundries,
or should we try a different approach?
Realizing that the market for metal parts is less risky and
much larger than the market for equipment aimed at the foundry
industry, we decided that, instead of trying to sell our
equipment, we would make metal parts directly for
concluded that our mission is to become a “one-stop shop for
Soligen was then
organized to become a unique manufacturer of metal parts to end
users, at any quantity, directly from a CAD file.
Our unique position in the market would be, unlike other
metal parts manufacturers, to offer the customer the metal parts
without the need to invest in patterns or molds prior to producing
and testing the first article.
Utilizing DSPC to initially produce the first articles
and then the patterns or the steel molds for mass production, all
the parts will be related directly to the original design, since
either the parts themselves or the tooling used to fabricate
them will be generated directly from the designer’s CAD--and
that’s exactly what Gutenberg’s printing revolution did to the
Perhaps we could
become “printing houses,” much as Gutenberg did.
In June 1994, we acquired our first foundry and machine
shop, and in January
1995 we opened a new business called Parts Now™.
Soon thereafter we experienced proof that Parts Now™
has merit. Caterpillar
challenged us to make complex engine components within a week.
It would have taken months to produce similar parts by
traditional methods. Parts Now™ dazzled Caterpillar as it
delivered fully functional parts within five business days from
receiving the CAD file over the modem.
I never checked
the Guinness book of records, but I believe that functional
testing of engine components, five working days after they were
taken from the drawing board is a new record in producing
functional metal parts. We’ve
proved, in less than three years, that the Parts Now™ concept
works. Thank goodness
I pursued my interest in metal parts in spite of hearing
skeptic’s comments. I
believe that the ability to order functional metal parts via telephone,
and to receive the first part in days, instead of months, will revolutionize
manufacturing and change the way engineers work.
The Caterpillar case is just an example of the flexibility,
and additional new degrees
of freedom in designing new products, provided to engineers in a manner
which is faster than any traditional fabrication process.
To me, these are
the early signs of a dream coming true. I'm not going to predict that Parts Now™ will affect the
human race as significantly as Gutenberg’s press, but who
knows how far our work will