align=”right” border=”0″>Business Week. The Economist. Magazine covers are glowing with nanotechnology stories-by some forecasts, the fastest-growing domain around, spanning semiconductors, health care, consumer products and even automobiles. Venture Capitalists and their limited partners have heard it for some time: Nanotech is hot.
When we measure things in nanometers, or one billionth of a meter, it means we are looking at individual molecules or even atoms-seriously tiny. The word “nano” is thought to derive from the Greek noun for dwarf. More recently however, it seems to derive from the verb that means “to seek and get venture capital funding.”
But the real opportunity may lie in another tiny technology called MEMS. It wasn’t too long ago that similar hype surrounded MEMS, or micro-electromechanical systems). Also playing to our fascination with things small, MEMS promised to use “micromachining” technology (much of it borrowed from the semiconductor industry) to fabricate miniature versions of familiar items. Pumps, sensors, relays, motors, just about any man-made gizmo in our physical world could be shrunk down to millimeter or micrometer scale, so that it could be hidden or worn or even implanted. By employing automated, parallel manufacturing techniques, MEMS technology would enjoy the same economies as integrated circuits, and hopefully ride similar cost curves.
But if nanotechnology is still in its infancy, MEMS is entering adolescence. Investors interested in making money in “things small” may want to look here at the more mature “micro” rather than the sexier “nano” as a way to make money in this cycle. Also, by looking at the successes-and failures-of MEMS, we can better understand what we might expect from its “little” brother. The siblings share some important attributes: Both rely on a progression of improvements in fabrication, test and measurement of things increasingly (or decreasingly) small. And both have been promoted as “horizontal” technologies with broad applicability. But perhaps that’s where we can find one of the key lessons-the successes and failures are more about how the technology is applied, than about the nuances of the technology itself.
Many investors cringe when they hear mention of MEMS. They’re probably remembering one or more of the ill-fated “optical MEMS” projects of the 1990s. Back then, technology using an array of dozens or hundreds of electrically-actuated mirrors, optical MEMS switches could obviate the need for back-and-forth conversion to electrical signals. The all-optical path could operate at any wavelength, any speed and any protocol. That’s why Nortel acquired Xros in March 2000, for stock valued (at the time) at $3.25 billion. While that was great news for Xros investors, it turned out badly for Nortel: They shuttered the effort within two years. Table 1 lists several startups that managed to exit during the height of the bubble, when there was some concern that optical networking would become so popular that critical components would be in short supply.
The Xros switch and others like it were technically impressive and even functional, but as the bottom dropped out of the telecommunications market, it seemed there just wasn’t much demand for an all-optical switch (no matter how capable or tiny). In a parallel universe in Dallas, Texas Instruments was working on a MEMS device consisting of an array of electrically actuated mirrors. Not just hundreds or thousands, but hundreds of thousands mounted on a single chip. Instead of switching signals among fiber optic cables, TI targeted its Digital Light Processing technology to video displays, and 5 million of them have been built into video projectors and big-screen televisions. Because it focused on the right application, TI’s Digital Micromirror Device turned out to be a big-scale success for a small-scale invention.
There have been some other notable MEMS successes. HP’s Inkjet printers employ a type of MEMS called microfluidics to precisely control the delivery of ink from an array of tiny orifices. Recent technology advances have upped the resolution, quality and efficiency of inkjet printers. Now, medical researchers are using inkjet print-heads to dispense not ink, but DNA and other proteins in ordered arrays for automated analysis.
Another high-volume MEMS component is at the heart of automobile airbag systems. A tiny MEMS-based accelerometer detects with great precision the sudden slowing of the vehicle and fires the airbag only when appropriate. Now produced in the millions every year, these devices have gotten smaller, cheaper and more capable, riding the learning curve originally promised by MEMS. And there are additional automotive applications emerging, including fuel pressure and air flow sensors, as well as collision avoidance and skid detection systems. Regulation drives much of the activity in transportation. Case in point: the Transportation Recall Enhancement, Accountability and Documentation (TREAD) Act of 2000 mandates that 10% of new automotive vehicles use tire-pressure sensors starting in November 2003, ramping up to 100% by October 2006. The total market for these tire-pressure monitoring systems will grow from approximately 4 million units in 2003 to more than 40 million units by the end of the decade. We’ve looked at projects marrying MEMS tire-pressure sensors to wireless networking systems in order to meet this safety mandate.
MEMS technology can be applied to real problems in the electronics world, so Radio Frequency MEMS (RF-MEMS) has captured our interest at Blueprint Ventures. Today’s RF systems employ many active electronic elements-transistors configured to amplify, modulate, demodulate, encode and decode signals as part of the transmitting and receiving process. As those electronics are integrated and shrunk with the progression of semiconductor technology, the “passive” devices used to switch, tune and filter the signals start to consume a disproportionate share of real estate, power and cost in radio-based systems.
RF MEMS addresses that challenge. As mobile devices grow in complexity, this technology allows them to support multiple frequency bands and new protocols without giving up performance, battery life or (the favorite feature in handsets) small size. There are a number of start-ups developing RF MEMS technology, including a Blueprint company in Irvine, Calif., called Wispry. Wispry doesn’t just borrow from IC-based manufacturing processes; it actually uses the same fabrication facilities and can build MEMS devices on the same substrate. Rather than a technology looking for a market, we know there are hundreds of millions of handsets being cranked out each year that will benefit from Wispry’s innovations.
Capital requirements for a MEMS production line can be large. And, as with semiconductors, the smaller the features, the more expensive it is. (Remember that a nanometer is one-one-thousandth of a micrometer, so watch your capital budget when you make a nanotech investment!) Many MEMS startups expended plenty of investor dollars wrestling with exotic materials and getting pilot production lines in place, exhausting their funds before their markets could develop. Innovative Micro Technology in Santa Barbara, Calif., provides “outsourced” MEMS fabrication services for a bevy of “fabless” MEMS innovators. IMT recently closed a $17 million investment round. Its facility has been re-purposed from making thin-film heads for the disk drive industry and has plenty of capacity to serve its clients as their “killer applications” take off.
As we continue to think small, we’re cautious about getting too far to the “R” side of R&D. We prefer to leave that to the biggest companies with the deepest pockets. But as MEMS and nanotechnologies mature, application-driven opportunities are emerging as are startups that are born to address them. MEMS may just lie a bit closer to the reality side of the capital efficiency frontier. With our model of capital efficiency, leveraging the investments of others as we put our dollars to work bringing innovative, cost-effective products to established markets may make better sense for MEMS than it does for its younger cousin. Profits, after all, should not be nano-sized.
Bart Schachter is a managing partner and David Frankel is a technology partner with Blueprint Ventures. Schachter focuses on comm. and IT infrastructure, wireless technologies, nanoelectronics, software and comm. semiconductors. His email is