Perspectives on StereoLithography
Automated Fabrication in the 19^th, 20^th, and 21^st Centuries

     When Guglielmo Marconi sat on the shores of Nova Scotia in 1901 and listened for the faint sounds of the Morse code for the letter “S” being transmitted by his laboratory assistant on the other side of the ocean, could he have imagined that less than a hundred years later his invention of the artificial manipulation of the electromagnetic field would be entertaining and informing billions of people around the world on a minute-to-minute, round-the-clock basis, not only connecting British citizens with the actions of their military in the South American Falkland Islands and Americans with their soldiers in Iraq, but even relaying photographs from a tiny Voyager spacecraft at the very edge of our Solar System?

     Like radio, automated fabrication is a radical step in the progress of technology, and you, Ladies and Gentlemen, Users of StereoLithography in its earliest stages, are to be congratulated for your foresight and your fortitude in seizing this opportunity and dealing with the host of challenges it throws you. And like radio, automated fabrication will grow into an industry, and become the technological basis of other industries, with its ultimate ramifications being far beyond our abilities to imagine. Tonight, I would like to offer a little perspective on our adventure by looking back at some of the historical roots of modern automated fabrication, then jump ahead into the future to see where all of this may take us, and finally come back to 1992 and the issues that must be faced to ensure measured and beneficial growth of this industry.

     Before I do that, let me pull back for a minute and explain this term that I am using, automated fabrication. Most of you are used to calling the technology “rapid prototyping,” so what is this “automated fabrication” that I’m talking about? As I see it, rapid prototyping is the first successful, commercial application of the automated fabrication technologies; it is not the be-all and end-all of the technologies themselves. There are, all together, five broad areas of application. They are:

• Industrial models and prototypes, the source of the name “rapid prototyping,”
• Patterns for molds and dies,
• 3-dimensional imaging of complex mathematical, chemical and biological data,
• Actual low-volume manufacturing production, and
• Artistic, computer-driven sculpture.

     One more diversion I will take before going on to talk about the past and the future is to comment on the importance of StereoLithography in this industry. Marconi was not the first person to think of using the electromagnetic field to transmit signals. There is even some debate as to whether he was the first to build a working transmitter and receiver. What is true without question, however, is that he was the one man who exploited the Dickens out of the technology and turned it into an industry, becoming one of the richest men in the world of his time in the process. It is in a similar way that StereoLithography is important in the sphere of automated fabrication. Chuck Hull was not the first to think of using computers to control the generation of 3-dimensional objects. He did not invent photopolymers. I’m not sure if he was the first to think of curing successive layers of photopolymers by a laser beam, although he does seem to have been the first to reduce that idea to practice. But one thing is certain. Chuck Hull took his idea of StereoLithography in 1982 or thereabouts, and built it, first into a laboratory curiosity, then into a machine, then, with the help of Ray Freed, into a company, and I daresay they built that company into the flagship of a veritable industry. Although Hull and Freed may hold stock only in 3D Systems and not in its competitors, they will forever be remembered as the founders of the whole automated fabrication industry.

The Past

     So where did the idea of automatic generation of 3-D shapes first show up? Let’s have a little fun with this. Who would like to venture a guess, or knows the answer? Just shout out the year that you think the first automated fabrication device was built and exhibited. … The earliest evidence I have found so far dates back, believe it or not, to 1717, when Peter the Great of Russia gave as a gift to the Institute of France a special lathe with a pantographic design for copying the relief of medallions. In 1804, James Watt, who was then retired from his epoch-making work as inventor of the steam engine, is reported to have seen this or a similar lathe on a visit to Paris. He returned to his home near Birmingham, Great Britain, inspired by this idea, and spent the remaining fifteen years of his life working on an ambitious variation which was capable of copying not only medallions, but fully sculptured human busts. His device actually worked, and Watt made copies of several well known sculptures of the day.

     Watt’s sculpturing machine worked by milling the copied shape out of a block of solid material, usually either wood or ivory. So you might complain that while it may have been the forerunner of automated machining, StereoLithography is a world apart from that because it is an additive process, instead of subtractive. Agreed. So let’s set aside subtractive processes and ask the question again. Who had the first idea for building up 3-dimensional shapes by the automatic processing of some kind of raw material? Anyone want to take a guess at this one? … Let me give you a hint. I will read from the second paragraph of the patent:

My invention relates to improvements in photographic processes for producing plastic imitations or representations of objects of all kinds; and the purpose of my improvements is to obtain a set of photographic plates or films which in superposition will so modify the light passing therethrough that all modulations of light other than those due to the differences in relief will be neutralized. If, therefore, such a set of plates be copied on suitably-prepared gelatin and subsequently treated with water or other suitable substance, an exact imitation in relief of the object photographed will be obtained.

Be it known that I, Carlo Baese, a subject of the King of Prussia, Emperor of Germany, and a resident of Berlin, in the Kingdom of Prussia, in the German Empire, have invented a new and useful Improvement in Photographic Processes for the Reproduction of Plastic Objects, of which the following is a full, clear, and exact description.

     The rest, as they say, is history. The gelatin referred to by Baese was a naturally occurring photopolymer in wide use at the time in photolithography, the old-fashioned, 2-dimensional kind of lithography. The use of photopolymers went through various gyrations, such as exposing through masks and exposing by intersecting laser beams, until the idea arose of scanning the surface with a laser.

     Outside of the field of photopolymers, other approaches were tried, most notably the work of Thyssen, the German steel maker, in shaped welding fabrication in the 1970s. Thyssen built huge pressure vessels weighing up to 79 tons by this method. Like the early photopolymer experiments, Thyssen’s work never saw commercial application, however, it is being reincarnated in process testing carried out recently at the University of Nottingham in Great Britain. In the 1980s, work began on the processes that today make up the primary commercial competition for photopolymer-based fabrication: laser sintering, robotic extrusion and cut-sheet bonding.

The Future

     With that encapsulated history of automated fabrication under our belts, what do we get if we turn our window of inspection in the opposite direction and gaze into the future? There are five likely areas in which we can expect to see development in this industry. They are:

• Materials,
• Processes, including improvements in speed and accuracy,
• Cost,
• Software, and
• Robotic interfaces.

     Materials. One of the things that has amazed me about this industry is how quickly it has responded to the need for versatility in the materials that can be fabricated. Even for the StereoLithography Apparatus, which is supposed to be limited to that single kind of material called photopolymers, the chemists have come up with an impressive array of materials properties, from a soft, flexible material that feels like a hard rubber to a tough, machineable plastic able to withstand a good deal of abuse. The ability to extend photoinitiated polymerization to a wide variety of monomers may eventually bring almost all of the diversity of plastics manufacturing into the domain of automated fabrication. And since some users, including some in this room, have already been experimenting with interrupting the SLA process to embed foreign materials in the uncured resin, we can expect to see those processes automated to yield advanced composite materials. Meanwhile, developments in the competing technologies will bring us to fabrication in metals and ceramics. Within fifteen years, it is feasible for the entire inventory of natural and man-made materials to be within the domain of automated fabrication, including soft organic tissues and refractory metals.

     Processes. As revolutionary as today’s fabricators are, chances are that anyone under the age of seventy will see the day when we will look back and chuckle at how slow and sloppy they were. The ultimate direction of this industry will merge its machines into the realm of nanotechnology, where objects are built up one molecule at a time. Machines will combine low- and high-resolution modes to allow the fast generation of bulk material, yet with intricate details expressed in selected interior and exterior regions. Chuck Hull predicted at last year’s Dayton conference that “the ultimate resolution of photopolymer laser printing is sub-micron.” In addition to today’s additive processes of photopolymerization, laser sintering and robotic extrusion and welding, we are seeing dramatic improvements in the old, standard, subtractive ones, giving us such processes as micromachining, microgrinding and microetching. Newer additive technologies are also coming along, such as chemical vapor deposition. And eventually, we will even see automation applied to the compressive processes; a first step in this direction is the work being conducted on electromagnetic forming.

     Cost. I am tired of hearing complaints about how expensive fabricators are. At a half-million dollars, the highest price tag in the industry, the one word I can think of to describe a machine that can automatically generate any arbitrary shape in a stable material to within a few thousandths of an inch of its design is “cheap.” The best analog we have for the advent of fabricators in the 1990s is that of computers in the 1950s. In each case, the new machines are high technology for their times, require highly skilled operators and special environmental controls, and offer such incredible capability that most people on the planet can’t even imagine what useful purpose they could serve. The main difference that I have noticed between the Sperry-Rand UNIVAC and the 3D Systems SLA is about an order of magnitude in price, probably two or three orders of magnitude if one were to account for inflation.

     With this in mind, I see the price of fabricators moving in two directions. We will see small, tabletop “personal factories” that will fabricate everything from guest-customized dinner plates to replacement dishwasher parts, and will cost, like personal computers, about the same as a car. At the other end of the spectrum, room-sized “superfabricators,” generating large volumes of biological-scale mechanisms, will fetch multiple millions.

     Software. We all know that the bottleneck holding up explosive growth of this industry is CAD software. Just as computers were brought down to Earth by the Macintosh, autofab will become a mainstream technology when it is as easy to create a 3-D design as it is to type a document into a modern word processing program. This will require an operating environment using virtual reality and real-time holographic projection, running on a neural network, which are all today still laboratory curiosities. Give it twenty years.

     Robotic interface. The last foreseeable area of development of automated fabrication is in the realm of what is done with the stuff that comes out of the machine. If a human being has to reach in and pull it out, then we are missing a tremendous opportunity to engage the fabricator as an element in an entire chain of automation leading from design through not only production of parts, but also assembly of working mechanisms. Peter Sferro has foreseen a day when customers will design their own unique cars at a terminal in a Ford showroom. This will require the unique aspects of these cars to be made on autofabricators, but to be practical, it will also call for the output of those fabricators to be automatically removed, combined with standardized components, and assembled.

The Present

     So now we’ve looked a few centuries into the past, and a few decades into the future. Let’s use the perspective gained from that investigation to come back and look at the most important period of history ever, the ever-present here-and-now.

     If history teaches us anything at all, it is that good ideas do die, so there is no guarantee that the future that I just described will actually take place in our lifetimes. The fact that the likes of Ciba-Geigy, Du Pont, Sony and BF Goodrich have invested a total of probably between $50 and $100 million into these technologies is a good sign that they won’t fade away overnight. But they could fade away if several key measures are not taken. If the discovery of a dramatic level of utility, as has been made by you enthusiastic SLA users, is not repeated and multiplied many times over by new customers of both 3D Systems and its competitors, then the whole industry will wither and die, leaving it to a future generation to reinvent.

     Let’s try an experiment here. How many SLA machines are represented in the room here tonight? About 100? 150? Let’s say I have a large sack of money here, about $100 million dollars. How many people would be willing, on behalf of your companies, to deed over your SLAs to me right here tonight for $1 million in cash, if I made you promise not to buy another one for a year? … You see, that demonstrates what I call a dramatic level of utility, that you are not willing to take double or quadruple your money to get out of your investment. That’s why you’re here. That’s why your companies are continuing their investment in the technology by sending you to this meeting, because your SLA has been paying for itself at a rate of probably between once and several times a year.

     Now, there are some people who say that the market for SLAs is saturated. As evidence, they point to 3D Systems’ sales for 1991 compared to 1990. Well, I have an advantage here when it comes to understanding this kind of situation, because I’ve ridden on this kind of roller coaster before. In 1982, I was selling IBM PC “clone” computers and I sold all I could make. A couple of years later, after I had got out of the business to pursue other interests, the market was glutted, sales were down, margins were down to a few points, and the industry trades were expounding the morose thought that all of the companies that needed desktop computers already had them. Does anyone here remember that?

     The problem with theories like that is, although they turned out to be total nonsense, it was difficult to know at the time whether to take them seriously. But if we use the benefit of hindsight, we can see that, in fact, there were hundreds of thousands of companies that did not have computers and could neither compete nor cooperate with the other companies that did unless they got one, or several. We also see that large companies that had installed hundreds of original PCs wanted to repeat their jump in productivity by moving up to 286s and then to 386s, but this time they installed thousands.

     In my new book, I list 18 categories of users and potential users of autofabricators. Copies of this list have been distributed to you at your seats. The first six categories have already exhibited some form of success in using these technologies. These are the manufacturers, industrial designers, surgeons, archeologists, sculptors and jewelers. The other twelve have not even begun to think about automated fabrication. These are the movie propmakers, architects, decorators, chemists, military strategists, etc., etc., etc. There are literally millions of professionals in the United States alone who could benefit from the ability to represent their designs or data in solid three dimensional form. Why aren’t they using autofab? Is it too expensive? Definitely not. Is CAD too difficult to learn? Yes, but the benefits are worth the effort. Is the accuracy too rough? Not for most applications. These people aren’t using automated fabrication because they don’t know about it. And I don’t mean they haven’t seen the clip on Good Morning America or read about it in Newsweek or the Wall Street Journal. They don’t know about what it can do for them in their businesses. That’s it.

     So what can be done about this? The solution to this problem will require three elements. The first is already being done, and that is a concerted sales effort. 3D Systems has its sales force in the field and every other new vendor is out there knocking on doors. Those of you users who are engaged in a service bureau business are doing the same. The word is spreading. The industry is so young that every new customer brings a unique set of problems that haven’t been experienced before, and every time that happens a new set of solutions is produced that have the potential to attract another dozen customers.

     Automated fabrication is an infant industry or, as Rich Binning puts it, an industry in gestation. When Marconi had proven that radio could be used for trans-Atlantic transmission, the nay-sayers scoffed that it would only replace one trans-Atlantic cable because only one radio signal could be transmitted at a time. Of course, Marconi responded by inventing frequency-tuned transmission. But the important lesson is that nay-sayers do not kill an industry. The only thing that kills an industry is the demise of enthusiasm among its vendors and customers. That 150 StereoLithography users have taken three days out of their work weeks to compare notes and offer feedback to their vendor speaks well for the future of this industry. So does the fact that 3D Systems is here with such a strong contingent to listen to and respond to your feedback. I look forward to hearing from some of you over the next few months as to quality of their response.

     The second element required to move this industry forward is a broad expansion of the range of applications to which its tools are put. To go back to the example of radio again, how large would the market for that technology have been if it had never been used for more than broadcasting Morse code? Well, we are in the Morse code days of automated fabrication. One example of an application waiting to be exploited is the one I call 3-D imaging. Look back at the list of users and potential users again. The column of check marks for 3-D imaging shows that a couple of daring individuals have ventured into this territory in the fields of medicine and archeology. But it is among the future users that this application will really begin to be appreciated for all it’s worth. Scientists will use fabricators to build models of molecules, cells and organs; engineering instructors will make 3-dimensional cut-away models of complex mechanisms from various perspectives to demonstrate their workings to students; forensic artists will forego the old, crude sketch of a suspect in favor of a lifelike, full head-and-shoulders bust; military tacticians will render the terrain of a battleground in 3-dimensional relief to truly understand the challenges to be faced by their troops.

     There is another important characteristic of many of these potential users of automated fabrication for 3-dimensional imaging. They have money. Biochemists needing to understand the mechanisms by which the AIDS virus attacks T-cells are some of the few scientists who can feel fairly secure in the continuity of their government funding. The current political excitement about “fighting crime” in the United States has rendered several local, state and federal agencies flush with funds available to explore new high-tech techniques. And, of course, Cold War or no Cold War, the world’s defense agencies will always be among the best-funded departments of their respective governments. Three-dimensional imaging is a vein of gold waiting to be mined. The smart autofab vendors and service bureaus will be out there with their pick-axes and shovels, leaving no stone unturned to make sure that everyone who both:

• Can use their machines for this purpose, and
• Can afford to either buy a machine or buy the services of a machine

     So the first two elements required to advance automated fabrication into the prominence it deserves are continued sales promotion and applications development. The third element is patience. It is difficult for people who have a glimpse of the future to give the rest of the world the courtesy of time to catch up to that vision. But that is exactly what is needed here. Automated fabrication will be a multi-billion dollar industry, but not next year, and not in 1995. It probably will be by the turn of the century, but only by slow, methodical effort on the part of all of the vendors, users and support industries. Those who enjoy a challenge will thrive in this environment; those who expect the world to make itself over in a day will be frustrated and burnt out.

     Ladies and gentlemen, I hope that your sessions here at the Users’ Conference have been productive, and that you will go back to your laboratories and factories and continue to enjoy the hard work of pioneering in the use of a new technology.

Handout

     If you found this interesting, you’ll also want to read:

• The Origins and Direction of the Fabricator Revolution, the farthest-reaching discussion of future fabbers, including “accretive fabrication,” modeled on the growth of biological systems.
• Automated Fabrication—The Freedom to Create, based on Marshall Burns’ first corporate keynote presentation, compares the impact of fabbers on manufacturing to the impact of books on literacy and of cars on transportation.