MEMS Ingenuity

I was at the International MEMS/MST Forum all day yesterday and the one fact that jumped out at me was that approximately 250 MEMS companies make less than $35 million dollars. This industry has a long tail.

MEMS is a technology that represents an enormous amount of creativity and manufacturing ingenuity. There is a spirit entreneurialism at work that reminds you of the early days of the PC and the dot.com era. People have ideas and dreams and skills and they believe they can make a difference. They believe they can be the first to see a need and work out the engineering and get on the cost curve and beat the big guys. The industry isn’t yet about scale, it’s about cleverness and taking on risk. While the market numbers are still dominated by the big LDP and printer apps from TI and HP, the MEMS business is still characterized by one-off devices where someone sees a better a way to marry a sensor with a microprocessor and do it better than companies with thousands of engineers and tons of capital.

The process of creating a device seems to be often one of basement creativity and engineering on-the-fly. First you have the application: integrating pressure, temperature, motion and other inputs with a control system for an automobile, a cell phone, a camera or one of the hundreds of other devices that interact with people and an analog world. Everyone who uses a cell phone or power tool or industrial device could probably think of ways to marry functions among parts or create new features or functionality. Then there is the real tough engineering of developing a faster-better-smaller-cheaper package with a dependability and robustness that can attract a big customer. Getting the first products to market seem to entail a whole lot of trial and error—devising new manufacturing methods and packaging techniques that need to be customized by the device, by the process and by the fab. Testing the part often requires that development of an entirely new testing protocol and specially designed equipment. You don’t do this stuff on computers and with formulas. You need to get your fingers dirty.

There will be a shakeout in the market and many of these small companies will be overwhelmed by economics, competition, fate, bad luck. And new processes, design tools, processing equipment and test systems will make the market more efficient, more standardized, more about scale and optimization. But while MEMS have been around for a while, it’s still a business characterized by dreamers and doers and hands-on engineering and that’s a good thing.