Manufacturing Tiny Parts One Layer at a Time
Conversations about innovation typically begin and end with megatrends: the invention of the microprocessor, the discovery of carbon nanotubes, Google's (GOOG) algorithms for Internet search. But some of the most important advances are born deep inside the manufacturing supply chain and are never seen by consumers, even though they transform products we use every day.
Arthur L. Chait, president and chief executive of EoPlex Technologies in Redwood City, Calif., is the architect of one such transformation. His company has developed radically new technology for fabricating pill-size electronic components for wireless communications, clean-energy applications, semiconductor packaging, and more. (See also the CEO Guide to 3D Computing, BusinessWeek.com, 10/6/2008.) Although very small, each component is a complex sandwich of ceramic, metal, or polymer layers that perform different functions. And EoPlex can crank out these parts—everything from tiny antennas for cell phones to key elements in pressure sensors for car tires—more quickly, in higher volume, and in many cases, at lower cost than standard manufacturing methods.
To see how this process works, it helps to understand how components such as cell-phone antennas are normally produced. Each phone may contain a half-dozen different antennas, which allow it to communicate simultaneously with a cell tower, a Bluetooth headset, a Wi-Fi network, and also receive GPS satellite signals. Today, the metal and ceramic elements that form the antenna are stamped from sheets of materials and assembled by robots.
Bake and Sinter the Substrate
EoPlex's technique more closely resembles printing semiconductors on silicon. And, just as in the semiconductor business, the shift in technology opens the door to improved designs at lower cost. As Chait describes it, the process lays down whole sheets of different materials in patterns, building anywhere from 20 to 500 layers. Then the resulting substrate is baked and sintered at high temperatures into finished components. Each part, Chait says, contains "all the necessary internal features, already in place." Depending on the application, these might include channels, chambers, sensors, and conductors—all exactly as the design specifies. "It still amazes me that it happens the way it does," he says.
The same process, Chait says, is ideal for fabricating thousands of so-called energy harvesters—tiny elements in pressure sensors used to keep car tires inflated at just the right level. Such sensors are now required by law on all new cars sold in the U.S. The typical candy-bar-size unit is battery-powered, and the device is sealed against moisture and road chemicals. When the battery becomes too weak to operate, the entire unit must be replaced at considerable cost and annoyance to the consumer. And the sealed unit must be recycled to avoid potential pollution.
EoPlex says it has a better idea. It has produced prototypes of energy harvesters that capture energy from tire vibrations, eliminating the battery completely. These devices contain tiny tuning-fork structures made from piezoelectric materials, which have the ability to convert mechanical energy into electricity. In the tire, they convert tire vibrations, and the electric current they generate is stored in a capacitor for about 20 seconds. At that point, there is enough electricity to power the whole device. Powered entirely by piezoelectrics, the pressure sensor sends a steady stream of signals to the car's onboard computer, which alerts the driver if the tire is under-inflated.
Ultimately, such devices will be cheaper, greener, and more durable than today's generation of pressure sensors—though they may take several years to reach the market.
Chait didn't invent EoPlex's process. The credit for that goes to Charles S. Taylor, an engineer and serial entrepreneur whose earlier startup, a surgical device company, was sold to Guidant in 1999. To help launch EoPlex in 2001, Taylor recruited Chait from contract manufacturing giant Solectron (now part of Flextronics) (FLEX), in Milpitas, Calif., where he was in charge of major accounts including Cisco (CSCO), IBM (IBM), Nortel (NT), and Dell (DELL). Chait, who has a bachelor's degree in ceramic and materials engineering from Rutgers and an MBA from the University of Pittsburgh, raised the capital for EoPlex and recruited its core team of engineers.
Right now, the privately held EoPlex is manufacturing small commercial volumes of antennas and other parts at its Silicon Valley head office. For financing, it depends on its venture capital backers: Draper Fisher Jurvetson, Labrador Ventures, ATA Ventures, and Draper Richards. But next month, the 21-employee company plans to open its first dedicated production plant nearby. Chait says EoPlex should be operating in the black by the end 2009, and has raised enough cash to see it through the next 18 months.
The key to its success so far has been finding applications that no other manufacturing system can deal with, or that would be too expensive to handle any other way. "There are lots of other techniques that make parts in a small form-factor, but typically they are limited to just one material," Chait explains. These approaches include micro-molding and semiconductor techniques that etch away tiny sections of silicon. "But if the part is highly complex, the machining of the mold can cost hundreds of thousands of dollars," he says. As the parts get smaller and more complicated, costs rise exponentially. The patented EoPlex process "is one of the few ways you can handle this," Chait says.