Microsystems a Major Research Area
With the ubiquity of handheld devices and the advent of the Internet of Things—networking everything from cars to light bulbs—microsystems, which bring together functionalities such as sensing, actuation, information processing, and communication at chip-scale, has become a major area of research in electrical and computer engineering.
“A revolution of innovation is poised to be unleashed in the coming years,” says Amit Lal, a professor at Cornell ECE. It is not only driven by the “gee whiz” appeal of adding functionality to devices that allow users to sense and respond to the surrounding environment. “To a significant extent, the GDP of our country is dependent on the productivity we get out of these products as we use the devices all the time to do work,” said Lal. “In the future, cell phones are going to continue to be even more capable because the information they provide is a central engine for creativity.”
The leading information technology research firms IBM and Google agree that by 2020 there will be approximately 27 billion devices connected to the Internet. Many of these devices do not yet exist, and each will require small, efficient, and reliable microsystems to keep them connected as they provide seamless benefits—ranging from health monitoring to security—based on voice and fingerprint recognition.
Like revolutions in other areas of computing, the increasing use of microsystems will require parallel innovation in more traditional areas of research. “Because you have so many things connected, you need technological revolutionary ideas to increase the speed at which you can process and store this information,” said Lal. “All these handhelds are going to generate so much data that it’s going to end up someplace and you need to process it all in large quantities.”
Smart phones, already a collection of microsystems—a camera, a microphone, a few accelerometers and gyroscopes—are obvious platforms, and the addition of new functionality to smart mobile and spatially-bound electronics is likely to further change people’s relationship with devices that already give them positive feedback in the form of “likes” on social networks such as Facebook.
“Much like pets provide psychological comfort and healthier living through undivided attention and unconditional affection, I think in 15 years, it’s very possible that the cell phone will become pet-like, becoming a source of unconditional dedication,” said Lal. Using motion body detection, personal devices could figure out your feelings. Your cell phone might take advantage of the integrated inertial sensors you might have on your body and see how you are reacting to things in terms of your physical posture. Your cell phone may take that data, process it, and it might say, ‘Why are you fidgety?’”
“The concept of what constitutes a microsystem has evolved since the 1970s, when the focus was on microcontrollers, microprocessors, and microelectromechanical systems or MEMS,” said Lal. Today the definition has broadened to include any chip with the ability to interact with the environment beyond bits going in and out.
“Today a microsystem is a chip in which there are transistors, there is some way of communicating with the world
other than using wires—with photons, sound, mechanical, or chemical signals,” said Lal. “The prototypical microsystem that the MEMS field is going after is a tiny robot that can sense, compute, and move intelligently with integrated power sources.” The technology that can implement these functions could form the basis for miniaturizing many systems to
For example, says Lal, you could put several different types of microscopes—scanning electron microscope, ultrasonic sonars, and terahertz radars—now all very large, on one chip. “You could imagine putting these in the fingertips of a latex glove and then the doctor wearing that glove. Wherever he touches, he can image what is happening,” he said. “One type of microscope by itself may not be sufficient to tell you what’s going on, but when you have them all, the synergistic output from all is more than the sum of what each microscope can provide individually, and you can see what it is conclusively.”
An atomic clock, another bulky technology, has already been miniaturized to chip size. “By using micromachining, you can shrink it down to about a cubic centimeter and it’s commercially available now. That’s an example of a successful microsystem,” he said. “This example adds merit to the mantra that anything is shrinkable.”
The challenges in miniaturization, however, can be formidable. “Sometimes making things smaller improves performance. It takes less energy to heat objects as they get smaller, such as heating a smaller blood sample for DNA analysis. But in many other cases, you have to invent new ways to improve performance at small scales, as exemplified by the large numbers of challenges facing further shrinking of the transistor.”
Cornell Engineering has long played a pivotal role in the area of microsystems.
Cornell graduates founded Kionix, one of the first and most successful MEMS manufacturers in the world. Kionix has guided the development of products for inertial sensing with a majority of manufacturing in Ithaca—proving that Ithaca can be a global manufacturing town. The success of Kionix in Ithaca has led to spinoff companies in optical microsystems (Calient) and microfluidic microsystems (Rheonix). Cornell has helped develop Ithaca as a powerhouse of MEMS R&D and industry.
One of the reasons Cornell Engineering was a natural place for MEMS research to take hold early on is the institutional commitment to collaboration across interdisciplinary fields. Another reason was the presence of the National Science Foundation-funded Cornell NanoScale Science & Technology Facility, which coupled with expert staff, supported a broad range of nanoscale science and technology projects by providing state-of-the-art resources.
“MEMS and microsystems tends to bring people together,” says Lal. He mentions the work of fellow Cornell ECE professors Michal Lipson, Rajit Manohar, Edwin Kan, Sunil Bhave, Farhan Rana, Alyssa Apsel, Ehsan Afshari, and Al Molnar. “Interdisciplinary colleagues are very important to MEMS because they provide expertise in optics, electromagnetics, circuits, algorithms, and complementary devices. Without advanced circuits and algorithms you can include all the sensors you want, but if you can’t make them work together, they are not very valuable,” says Lal. Not only does each component need to work in miniaturized form, but they also must be integrated so they don’t interfere with each other. And they need to be designed for production at high volumes to be affordable.
“Foundries make money by maximizing the number of wafers they process per week for a paying customer,” said Lal. “For integrated circuit (IC) foundries to make money they need to push more wafers through, and continued increase in microsystem demand could be crucial to continued profitability in the semiconductor industry.”
Currently, only a handful of foundries produce most of the silicon chips used in the world. And they work hard to reduce cost with competition and offer higher value with higher capability and process differentiation.
As the technology becomes less expensive, it’s possible to imagine silicon becoming as pervasive as steel or brick. “It’s not so crazy to think of everything made of silicon circuits—everything as in our architectural structures and automobiles. Especially if we learn to make silicon stronger than steel,” said Lal. “If you have a thousand foundries pumping out silicon at extremely low cost, you can start functionalizing everything around you with electronics.”
Lal does not know what the future of microsystems holds, but he knows it is wide open. “Have I thought about all the possible applications? No,” says Lal. “But that is what’s driving the future of electrical engineering—the idea that the more functionality we can put on the chip, the more likely that vision of functionalizing everything might come true.” In his own lab, this sense of the future being wide open is highlighted by the many lines of research Lal pursues. He directs the SonicMEMS Laboratory at Cornell ECE, which works on topics such as linear and nonlinear effects of ultrasound for microfluidics, applications of radioactive thin films for powering the “Internet of Things.” His group has also developed radioisotope-powered large-area nano-lithography. The SonicMEMS Lab has also developed ultrasonic microprobes for surgery and bioinstrumentation, nanoelectromechanical analog switches and computation, hybrid insect cyborg microsystems, micromechanical solar energy, and chip-scale high-energy particle accelerators.
Whatever the future of microsystems holds, research and innovation done at Cornell Engineering will play a big role.
by Jackie Swift Around 50 years ago, Gordon Moore, chief executive officer of Intel, predicted that every two years the number of transistors a microchip could contain would double. Known as Moore’s... Read more about Keeping Computer Hardware Fast and Furious