by Ping Fu
—
The Chinese have a saying: “Puppies don’t know yet to be afraid of tigers.” In English we say, “You aren’t afraid of what you don’t know.” I moved forward with the founding of Geomagic in a manner that may have seemed courageous to some. In truth, I was more like an innocent pup stumbling around a tiger’s den in the dark: I had little sense of the risks I was taking.
As soon as I left NCSA, I realized how much I had been taking for granted. All at once, I had nothing: no office, no phone, no computer, no lawyers, no accountant, and no money. Herbert and I put in some of our own savings to get things started.
Our first order of business was to find a suitable name before we incorporated. I went to the library and checked out a book about naming and trademarking. It said that either you could make up a word—such as Xerox, Kodak, or Kinko’s—or you could name your company by two common words that had nothing to do with each other. Apple Computer was a prime example of the latter: apples have zero connection to computers.
It was a rainy evening in Champaign, and the wind smashed against our windows as Herbert and I lay in bed dreaming up names. On the wall across from our bed hung a painting of colorful raindrops by Austrian artist Friedensreich Hundertwasser.
“How about Raindrop Geometrics,” Herbert suggested.
“I love it,” I responded with excitement. “Raindrops are points in space, and they form interesting shapes when they meet. This is a poetic version of what we want to do.”
The corporate attorney who helped us to register the company didn’t like “Geometrics”; he thought it sounded too technical. We settled on Raindrop Geomagic, and incorporated the company under that name. Years later, we dropped “Raindrop” because it was too long a name and people would end up using the abbreviation RGI, which I disliked. We couldn’t get “raindrop” as the domain name, but we had secured geomagic.com. It proved confusing for trade shows and company listings because people were not sure whether to look us up under R or G. Naming a company well is tricky, I learned, but not nearly so complicated as running one.
At the start, it seemed that Herbert and I were the perfect partners to found the company together. He had depth and I had breadth. He was the mastermind behind the technology and I focused on developing the applications and markets. He could stay in academia and bring home a regular paycheck while I risked everything to start the company.
Although we used Herbert’s mathematical insights and elegant algorithms to build our software at the beginning, it soon became clear that he wasn’t interested in the business issues. Solving real-world problems was too messy for his academic research. He was a theoretician who preferred spending his time contemplating elegant proofs of unsolved problems while his head floated above the clouds. He declined working for the company, but remained involved with Geomagic as an adviser. He never questioned my judgment or attempted to alter my decisions. When things got tough, he always offered his wisdom and support with an air of detached collectedness.
With our own savings as the initial cash investment, I hired two MBA students and naively asked them for a business plan in two weeks. They looked at me as though I had four heads. “Two weeks?” they gasped. It took three months and stock options for them to deliver. I enrolled in an entrepreneurship class—clearly, I needed to learn business management skills on my own.
Next, I hired two technical students for software development—Mike Facello and Dmitry Nekhayev, who remain at Geomagic today—and rented an affordable office space. It was an incubator in a manure field the agriculture department was using to experiment with human fertilizer. Every time I walked in the door, the putrid smell reminded me of the open latrines from my childhood in the dorms at NUAA. I laughed to myself: I had come so far, yet here I was again, literally in a pile of shit.
I marvel sometimes at the magic of the software our talented team created. The process started with digitally capturing a real-world object using a 3D scanner. The dots—also called a “point cloud”—gathered from the scanner were not projected onto a flat plane, but rather floated in virtual space. Geomagic software then magically linked the dots together into a coherent 3D digital model representing the shape of the object. Our algorithms allowed us to reach inside every nook and cranny of any object, no matter its size or complexity.
It is difficult to convey the mathematical power behind the software. Imagine making a model of a car from beads. The software’s job was to figure out how to string the beads—of a point cloud—together with a single thread, without ever crossing back over. The resulting model not only retained the overall shape of the object, but also recognized its functional components, such as doors, wheels, and body. All this could be computed in a matter of minutes. For many customers, we had reduced the process of scanning a physical object and creating a digital model from weeks to minutes.
—
We didn’t know how much money was needed to run the company, but bills kept arriving more quickly than Herbert’s University of Illinois paychecks. Our first employees courageously offered to chip in some of their savings, but it wasn’t enough. I applied for a bank loan. The loan officer took one step into our stinky incubator office, gave a sniff, and walked out with her nose held high.
Hong and her husband, Anselm Bischoff, had a popular and successful specialty retail store, Bischoff’s Shades of the West in old-town Scottsdale. Hong was excited about my new venture. She and Anselm offered me a great deal of practical advice and, when we needed it, cash. They were Geomagic’s first investors. My younger sister’s support meant the world to me—and it did to her as well. For the first time in our lives, I was following in Hong’s footsteps and she was able to help me, instead of the other way around.
Development costs came in higher than we’d expected. I learned that it took far more people and time to develop a commercial software product than an academic prototype. We hired more developers: Tobi Gloth came to us from Germany; Yates Fletcher, a brilliant mathematician, joined us from the Sun Microsystems graphics group; Thomas Jensen had valuable experience in medical imaging; and Steve Perkins became the technical support director with the ability to make every customer fall in love with Geomagic. I am indebted to these early employees for their passion, brilliance, and loyalty. In addition, there were lawyers’ fees for patents and trademarks, incorporating, taxes, liability, and all the other business terminology I was just beginning to learn about in my entrepreneurship class. We needed to raise money fast, this time from seasoned angel investors.
I had no idea that it would be so easy.
“Imagine your mother’s face as she receives a 3D print of her grandson’s first sculpture project,” I told a room full of angel investors and a panel of venture capitalists at the start of my ten-minute pitch at a Chicago investment conference in 1997.
“Now imagine viewing a 3D animation of your daughter’s teeth as they move into a perfect smile—on the day of her first visit to the orthodontist . . .
“Imagine walking into a shoe store, getting your foot scanned, and returning the next day to pick up your custom-fitted hiking boots . . .
“Now imagine your orthopedic surgeon printing a 3D model of your prosthetic knee, a perfect fit, a week before your surgery . . .
“While these examples may seem like dreams, Geomagic will transform them into reality.”
After my presentation, people bounced around me like chattering balls of energy. Each one had a story to share: “My left foot is bigger than my right, so shoes never fit me”; “I hate going to the dentist, but this technology would change that”; “Have you ever thought about using your 3D processors for . . . ?” Ideas tumbled into the room. It seemed that we had tapped into an unmet market need.
Donald Kurasch, an attorney and private investor from Chicago, handed me a check on the spot to make sure he got in on the first round. Peter Fuss, a venture capitalist and the former president
of Tellabs, followed me out the door and engaged me in a long conversation about our technology and its opportunities. He later became an investor and board member. Within a month, we had closed a $1 million round of private placement funding.
I felt happy. There was a deep current of humanity in our vision that resonated with people. I couldn’t believe strangers would give us money so easily. At the same time, I felt an enormous sense of responsibility and a little bit of uneasiness. Although I knew why I had started the company and what we could offer—mass customization software for the benefit of humanity—I had little clue as to how we were going to deliver on our promises or pay back the money and trust given to us by our earliest investors.
But never mind—our employees and I felt much more excitement than fear. Geomagic was a real company now.
—
My daughter was four years old when Geomagic got off the ground in 1997. I may have had my hands full running a tech start-up, but no one could get my attention better than Xixi. She was at an age where she wanted engagement, and she knew how to get it. Her smile started my day, and her hugs melted my worries away.
“I love having Xixi,” Herbert teased me. “I get to enjoy fantastic home-cooked dinners because she loves your cooking.” I laughed at his indirect encouragement. The fact of the matter is, if Xixi hadn’t demanded my full awareness, our family life and health would have suffered. It was truly a blessing to have a child so young when I started Geomagic: she was my priority, and Herbert and I both benefited from that. Without her, I would have worked every minute of my waking hours, including weekends. Being a mother also gave me deep respect for and understanding of other employees who already had families or wanted to start one.
Taking Xixi to the playground was a pleasure for me. While other moms sat on the park benches chatting with one another, I climbed onto the equipment with Xixi and played alongside her. One day, when it was raining, I took Xixi to an indoor playground. As usual, I tried all the activities: I bounced with Xixi in the padded jumping room, pulled out toys and tested each one, and scrambled like a monkey onto the climbing gym. Then, when I went down the covered spiral slide, I got stuck. The turn was too small for me. I called for help. A few kids came sliding down behind me and pushed at my back with their legs, but it wasn’t enough to dislodge me. Finally, two of the mothers came over. They reached inside to grab hold of my feet and yanked me out onto the cushioned floor. We all collapsed into hysterical laughter.
“You’re like a kid yourself,” one of the women observed once we had caught our breaths. It struck me just how true her statement was. Half of my childhood had been stolen away from me. Now I was getting to re-create it with my own daughter. I knew Xixi loved my involvement, but all at once I realized that my behavior was doing just as much for my well-being as it was for hers.
—
I made many mistakes in the early years of Geomagic, some of which were quite funny. Running a start-up is not for the faint of heart. I love how Reid Hoffman, the founder of LinkedIn, describes it: “You jump off a cliff and assemble an airplane on the way down.”
In 1998, a year after we got Geomagic started, we demonstrated our technology at SIGGRAPH, a premier computer graphics conference. At our booth, I met Mark Keenan, an Intel executive who invited me to present at headquarters. He said that he was intrigued by our technology because it pushed the limits of the central processing unit, or CPU.
“Should I bring my computer?” I asked. I wanted to feel confident that the software would perform smoothly when I demonstrated it. Also, I knew of Intel as a chip manufacturer; perhaps they didn’t have computers everywhere in their office, I thought. Mark said yes.
As I was preparing to leave, I realized that I had not thought about how on earth I was going to get my heavy desktop computer and monitor from Illinois to California. We didn’t have laptops to demo our software in those days; they weren’t powerful enough machines.
I bought a padded suitcase, put my desktop computer and a small monitor inside, and hauled it with me to the airport, where I checked it along with the rest of my luggage to San Jose. By the time I arrived, the suitcase wheels had broken off. Luckily, the computer and monitor had not been damaged.
I was staying with my friend David Knapp, the founder of a microchip design company in San Jose. He was surprised when I showed up with a large broken suitcase and a forlorn look on my face.
“What are you doing?” he asked.
“I’m going to Intel to give a demo, and they said I needed to bring a computer,” I said.
David laughed. “I kind of doubt it,” he said under his breath. But he was such a good friend that he figured out a solution. True to his maker’s nature, David had a wood and metal workshop in his garage. Stealing the plastic wheels off his large garbage bin, he fashioned me a rudimentary wooden trolley. The next morning we loaded the trolley, computer, and monitor into the trunk of my rental car, and off I drove to Intel.
I pulled up in front of the high-tech giant’s sparkling headquarters a half hour later. As soon as I stepped into the lobby, I wanted to run away. Here I was at one of the most advanced computer companies in the world, pulling behind me a squeaky homemade trolley. As I made my way from the front door to the reception desk, I could see people’s jaws dropping.
The receptionist called my contact, Mark, down to the front desk. Just as David had, he tried his best not to laugh out loud when he saw me.
“What happened?” Mark asked.
“My suitcase broke, so my friend made me this trolley to carry the computer.”
“Why did you bring it?”
“You told me to.” I started to giggle.
He winked. “I thought you were going to bring your laptop.”
Several people stopped and looked at my trolley in amazement.
“Well, did you bring the software?” Mark continued.
“Yes.”
“Does it work on a PC?”
“Of course,” I replied.
“Then I’ve got you covered,” Mark said. He had the receptionist store the trolley and suitcase with my computer in the closet behind the desk. We went upstairs carrying only a CD.
The software installed perfectly on the state-of-the-art desktop in Intel’s conference room, although I think my trolley left more of an impression than did Geomagic’s technology. Today, I wish I had kept the trolley as a memento, but David disposed of it after I returned to his house. He needed the wheels back for his garbage bin.
—
In 1999, we identified two potential launch partners with different manufacturing challenges: Boeing and Mattel. Boeing’s airplane parts were large, technically demanding, and difficult to replace. Mattel’s toys included an enormous variety of shapes, from Hot Wheels to jungle animals. Both could benefit from 3D imaging and printing technology.
As soon as the beta version of our software was ready, I made trips to both Boeing and Mattel to do demos. This time I knew better than to bring a computer along; we made sure the software had been preinstalled on the companies’ machines prior to my arrival.
I traveled to Mattel first. My knees shook as my contact introduced me to the twenty men gathered in the room: VP of manufacturing, VP of information technology, VP of products, and so on. I hadn’t expected such a high-powered group; I’d thought I was meeting a handful of engineers. A Mattel engineer loaded the data for one of their toy horses, which he had scanned. Within five minutes, Geomagic software had threaded the numbers together to create a complete 3D digital model of the horse on their computer. A designer could rotate it to view every angle, zoom in to see small details, and modify the design before sending it to production. I could tell they were excited because they all began speaking at once, talking about how useful this technology would be. A feeling of pride washed over me as it had when I’d turned on the lights for the workers at the Chinese factory decad
es before. It was a magical moment to have these people, whom I respected greatly, marvel at Geomagic technology, which our team had created.
I set off for Boeing with some degree of confidence. Their facility was located outside of Seattle. When I stepped into the meeting room, the audience appeared far less intimidating than the one at Mattel. There were only four engineers present, all with PhDs in mathematics. They wore jeans and gentle expressions on their faces.
“What do you have to show us?” one of them asked. I told them about Geomagic software. One of the scientists gave me a disc with the data from a 3D scan of a damaged door that had come off a Boeing 747 jet. I popped it into their computer and, just as it had with the toy horse data, our software worked its magic. I showed the engineers the 3D replica of the door from every different angle. Unlike at Mattel, I also attempted to explain some of the computational geometry algorithms behind the software, which I had learned from Herbert. I figured that, as mathematicians themselves, they would appreciate the insider knowledge.
A silver-haired man in a crisp collared shirt interrupted me. “I’d like to try something,” he said, showing me on his laptop the image of a cube with a cylinder inside of it. The cube had only corner points and edges like an outline, no walls. The cylinder was composed of just a few simple lines as well. “Let’s try to wrap this,” he said, loading the data onto the computer with Geomagic’s software installed on it.
It seemed like it ought to have been an easy enough task, far simpler than replicating a toy horse or an airplane door. But the Geomagic software failed to reproduce the cube and cylinder correctly. When I saw the messy lines that appeared on the screen, I felt like melting into the floor from embarrassment. All my childhood insecurities flashed in front of my eyes: I was onstage, declaring that I was a nobody. I’d taught myself math only a few years ago. How had I dared to think that I could make software to withstand the test of an advanced math group at an iconic American company?