Will You Soon Have Incredibly Small Computers Inside You? The Answer is Definitely Yes

And before you say it: You’ll be glad you do.

Groundbreaking technologies – and groundbreaking companies – often seem to come out of nowhere. We often have no sense of what the next big thing will be until the next big thing is already upon us.

So I thought I’d go behind the scenes at several companies building the technologies – and industries – of tomorrow and see how that works: not just the science, not just the technology, but the process, the decisions, the goals, the iterations… not just the nuts and bolts challenges of creating tomorrow but also the leadership and management challenges.

To find those companies I collaborated with and got support from Intel Capital, the global investment organization of Intel Corporation. They made finding the right companies easy: figuring Intel is much better at predicting than me at predicting what will be next, I simply chose five technology startups Intel Capital recently invested in.

This time I talked with Zhiyoong Foo, the CEO of CubeWorks, a Michigan-based startup building millimeter-scale computing with its Cubisens? platform, an autonomous wireless sensing platform with systems that can sense and process their environment, transmit results wirelessly, or store them for later use.

Oh yeah: the sensors are about the size of a grain of rice.

As happens with many startups, you didn’t set out to make incredibly small computers.

Our initial goal was fairly broad: we hoped to find commercial applications for some of the research coming out of the University of Michigan.

Our team is very good at ultra-low power circuit design, and we realized we are able to build all the building blocks that go inside a computing system: processor, memory, harvesting, power management, communication… so why don’t we try to build a complete system? With our latest iterations we’ve been able to build a complete pressure sensor, an autonomous wireless system, that is the size of a grain of rice. (Go here and scroll down and you’ll see a sensor resting on the edge of a nickel.)

How did that come out of being good at low power design?

What drives the size of most devices is the battery. A large portion of most devices is the battery; battery technology hasn’t evolved as quickly. So we decided that instead of developing smaller batteries we could reduce the amount of energy a device needed… which would then allow for a smaller battery and a smaller sized device.

For example, hearing aid batteries are some of the smallest “coin cell” batteries, and using those batteries our devices would last more than a decade. We’re basically below the energy consumption of a battery that dies sitting on a shelf.

So you know you can make small devices… but how did you decide what to work on?

That’s a question we struggled with for quite a while.

We believe in Moore’s Law but we ascribe to Gordon Bell’s law: there is a new computing class developed every decade. In the early days a computer was the size of a room. Then came work stations, then PCs, then phones… and we feel extremely small dimension computers is the next generation. Think of smart dots that can be scattered everywhere to perform all sorts of functions.

But that means in marketing terms you’re doing a push, not a pull.

Exactly. We’re engaged in a technology push, not a market pull. In large part the market doesn’t know they need what we can do – yet.

So one of the things we did was take a look at who needs devices this small, or smaller? Who has a pain we can solve? The first answer was the medical industry and the defense industry.

Medical applications are definitely interesting. Currently a number of solutions require you to go to the hospital and get a test. One example is pressure monitoring for kids with cerebral spinal fluid genetic disorder, or hydrocephalus, where there is excessive cerebral spinal fluid (CSF) in the brain. Your body produces CSF, and if it doesn’t get absorbed properly, your head grows. In extreme cases they’ve been drilling small holes to drain excess fluid.

So we talked to doctors and said we have a device that measures pressure that is incredibly small, and it appears that would be game-changing for them. Now you could implant the devices inside the shunt and the patient could leave the hospital and the pressure could still be logged. You’d get a lot more data, more information you could analyze… and instead of the patient having a headache and needing to go back to the hospital for tests, their condition could be monitored on an ongoing basis.

And on the defense side, it’s literally James Bond stuff. Unobtrusive devices, easily placed and transferred… there are all sorts of applications.

What our technology can do is allow for ubiquitous computer monitoring in situations where that is important.

There’s a paradigm that engineers are good at building things but not so great at deciding how they should be used or what people really need. In manufacturing we used to joke that we would tell an engineer what we needed and he would come back and tell us what we really needed… which rarely was.

I’m an electrical engineer. The members of our team have technical and research backgrounds. We’ve had to leave the comfort of our building and figure out where the possible applications are, whether our ideas are crazy, whether it’s the right market, how long before we should pivot away from a certain approach… it’s a learning process. I’m a fresh grad.

That’s one reason we took funding. We didn’t sell ourselves along the way. We actually turned people down.  But we couldn’t turn Intel down.

To us, funding is less about capital and more about tapping into their expertise and knowledge and network. When I went to the Global Summit I almost ran out of business cards. The leverage we can get from their network will help us work through the market learning curve we know we face.

Technical skills are clearly important, but how have you learned the leadership aspects of your job?

Starting a company is definitely correlated with personality.

I have a number of friends who are also doing startups. People that start companies are the kind of who are good technically but who also really like meeting people, talking, networking, learning from other people… and leadership skills can be learned.

We are fortunate that many of the professors at EECS (Electrical Engineering and Computer Science) at the University of Michigan are extremely good managers. They know how to raise funding, how to manage thirty or more PhD students who have very different goals and personalities, how to still manage to have a home life…it’s really impressive. You watch and learn.

People who start companies love to watch and learn. That’s one of the fun things about starting a company – you get to keep learning.

Say we project out five years. What do you hope will have happened?

My personal goal for this technology is to fulfill the dream of seeing our devices implanted in people in ways that make their lives better.

For example, people are working to close the loop for diabetics by creating injectable glucose monitors. What if the system was completely inside you? What if the system automatically gave you insulin when you needed it?  There is existing technology but it’s half in the body, half out; there are rejection problems, battery life is a problem… our stuff could be a game-changer. We could really change people’s lives.

That’s what I want to do. A lot of times we develop things just because we can, but I was fortunate enough to land in a group where the technology has real applications and can change the way we live.

As a society we will move there eventually. We are marching towards that, and we all feel we’re lucky to be a part of that.