When I interview researchers about their projects, I usually ask for a product timeline. It’s not an entirely fair question because researchers rarely have much control over commercialization. It’s a boring question too because they usually answer, “three to five years.” Still, I like to include these estimates in my stories to keep readers, many of whom aren’t familiar with R&D and product timelines, from getting too excited about the prospect of something like a quantum computer appearing under the Christmas tree.
Given the gap between research and commercialization, I’m excited to give an update on research I first covered in early 2006. A startup called MC10, based in Waltham, MA, will be making products from stretchable, single-crystalline silicon, the high-quality stuff that’s used in computer chips, within the next 18 months.
The great thing about stretchable single-crystalline silicon is that it can do anything chips can do now–sensing, processing, communicating–but it can do it in unusual shapes. Imagine a surgeon’s gloves that read a patient’s pH, a sheet of high-quality electrodes that can conform around the brain of a person with epilepsy, and a camera chip in the shape of an eye. Stretchable silicon can make these things happen.
There are other sorts of flexible electronics, but they are generally made of organic materials, which are printed or painted on flexible substrates. Organic electronics are easy to make, but aren’t as fast as silicon electronics.
MC10 is commercializing work from John Rogers’ group* at the University of Illinois. I wrote a story for Technology Review about the proof of principle: silicon becomes stretchy when it is thinned and cut into ribbons. These ribbons are then adhered to a rubber-like substrate that’s pre-stretched. When the rubber relaxes, the silicon forms waves, but does not crack. The rubber can be stretched and relaxed many times without degrading the silicon. Rogers showed the electrical properties of this wavy silicon are just as good as silicon on a rigid substrate.
Since this initial research, the group has made a number of working devices that can conform to different form factors and unusual shapes. David Icke, the head of MC10, gave a presentation at Printed Electronics 2009 on Thursday in which he provided an overview of stretchable silicon applications.
One project that the company will focus on is making skin for robots that perceives distance and senses pressure. Computer vision from cameras can get a robot hand only so far, Icke said. When it gets close to an object, it needs to adjust at a fine scale. Likewise, when a robot hand grabs an object, it needs to know the pressure it’s applying at all contact points. No one wants to be crushed between robot fingers.
After the talk, I caught up with Icke to ask exactly how they are adding perception to robot hands**. He says they’re building infrared emitters and sensors into a stretchable sheet that can be stretched over a robotic hand. The sensors pick up infrared light that’s scattered when it hits an object.
The early products, Icke says, will most likely be high-margin and low-volume. In other words, the company will focus on specialized gadgets, built for a specific customer, like spherical cameras for the U.S. military. So even though stretchable silicon products are on their way, it’ll be a while before you and I can get our hands on them.
*Rogers is a really nice guy, and one of the better science communicators I’ve ever come across. Rightfully, he was awarded a MacArthur Fellowship this year.
**Another way to add perception to a robot hand is being explored at Intel by Josh Smith. He’s built a robotic hand that detects an electric field, which is useful when in proximity to conducting materials like bottles of water or a human body. See a video here.