Computing

Breaking Boundaries

The Department of Electrical Engineering and Computer Science sees interdisciplinary research as its future.

When it comes to departments of electrical engineering, MIT ranks number one, according to U.S. News and World Report. It’s a fitting achievement for the school that was first in the United States to introduce electrical engineering classes and since then has been the birthplace of developments such as strobe lights and Rivest, Shamir, and Adleman-or RSA-public key encryption, the world’s most pervasive encryption system.

Today the Department of Electrical Engineering and Computer Science is a behemoth of some 120 faculty members and 2,000 students. They work and study in the shadows of such giants as Vannevar Bush, who developed the first useful computational machine, and Marvin Minsky, a pioneer of artificial intelligence who built some of the first mechanical hands.

But the department is never overshadowed by its past or overwhelmed by its present reputation. As it celebrates its centennial this month, the department continues to look to a future of breaking boundaries in research and teaching-from working across disciplines to using technology to overhaul the way courses are taught.

Life-Linked Research

Like the rest of the Institute, Course VI is awash in interdisciplinary research. The most prominent interactions involve work on biological projects.

“This century is the century of biology, like the previous century was the century of physics,” says associate professor Rahul Sarpeshkar, whose research group is working on a number of biology-focused projects. One is the development of a processor for a bionic ear. Cochlear implants can be connected directly to the auditory nerve, making hearing a reality for those profoundly deaf who still have the auditory nerve intact. Sarpeshkar’s group is creating a very low-power analog processor to interpret sound signals. Because of its extremely low power consumption, it will function for decades once it has been implanted inside a person’s ear. And because the silicon implant mimics the ear’s natural cochlear structure, it will be better than conventional hearing aids at distinguishing sounds amid irrelevant background noise. Within the next year or two, Sarpeshkar says his processor will be ready for use, at which time “you won’t even know the person is deaf.”

At the same time, the group is taking cues from nature in its development of other systems. For example, Sarpeshkar draws on the brain’s neural activity and left-brain and right-brain tendencies to inform his work on a hybrid computer, a machine that uses both analog and digital processes to compute. Also, to develop motion chips that, in a few years, could be used for target tracking, security cameras, and robotics, Sarpeshkar is taking lessons from houseflies, whose eyes are naturally highly sensitive to motion.

But this is not the only biology-associated project having an impact. Professor Eric Grimson, associate director of the Artificial Intelligence Lab, has been working in conjunction with physicians at Boston’s Brigham and Women’s Hospital on image-guided surgical processes. His computer systems use a patient’s preoperative scans to build a precise graphical model of the surgical area. Before surgery, doctors study the model to plan the least invasive means of completing their tasks. In the operating room, they project the graphical model onto the patient’s body to help them navigate. Also, throughout the surgery, Grimson’s systems track the surgical tools, showing the doctors the exact location of the tip of each tool and allowing them to guide it very accurately to the key structures they want to reach.

“The reason surgeons like this is, typically, it reduces surgery times by half,” Grimson says. “It lets the surgeons do surgeries they would otherwise treat as inoperable.”

Grimson’s system is used only at Brigham and Women’s Hospital, but he notes that similar-though less sophisticated-systems have started to appear on the market. He expects that in two to three years, pending government approval, such systems will be widespread.

Other notable projects include the research of professor David Gifford and associate professor Tommi Jaakkola. Their work links computer science with research on the human genome. Also, professor Jim Fujimoto has pioneered a new field of study-optical coherence tomography-that focuses on diagnostic surveys of the retina. And assistant professor Vladimir Bulovic, who develops devices that use organic materials as semiconductors, has helped produce organically powered crystals that glow in a variety of colors. These crystals could be used to make computer monitors that would consume much less power than today’s models.

“Electrical engineering, as one of the most mature engineering fields, has a lot to contribute to biology in terms of how you think about and approach problems,” says assistant professor Joel Voldman, who, along with assistant professor Jongyoon Han, works on biological microelectronic mechanical systems. Voldman has created an electronic method for holding cells in place so they can be studied.

In the end, department head John Guttag sums up his department’s growing focus: “You look at the department today, and it’s much more involved in both biology and medicine than ever before. we’ll continue to evolve in that direction.”

Tech-Linked Teaching

The department’s strong emphasis on teaching undergraduates is one characteristic that sets it apart from other institutions, says professor Jeffrey Shapiro, director of the Research Laboratory of Electronics. And these days, faculty members are being noticed for their particular use of technology to help them teach more effectively.

For the past few years, Grimson and his colleague professor Tomas Lozano-Perez have been experimenting with an online tutor for three courses. Their students can complete problem sets and submit their answers via the Web for immediate feedback. The online tutor lists the problems that were incorrect and gives hints on how to fix them. Students can work the problems again and again, resubmitting their sets as often as they like. “I like the fact that this turns the problem sets into learning opportunities rather than grading opportunities, and the students seem to appreciate that as well,” says Lozano-Perez, who plans to use the tutor in additional courses in coming months.

The pair also have been experimenting in 6.001, Structure and Interpretation of Computer Problems, with another idea-online lectures. In the past, the classes of 200 to 400 students were taught in part through large lectures. By the end of the term, Grimson notes, attendance usually had dropped to about 60 percent. So Grimson and Lozano-Perez eliminated most live lecture sessions and moved the lecture series online. Now students view audio-annotated PowerPoint presentations at their leisure, and most do so more than once. “Experimental studies that we have been conducting show that the students learn the material better, or at least as well, with online lectures than with corresponding live lectures,” says Lozano-Perez. However, both professors note that online lectures may not be the best option for other courses.

The department has taken another novel approach: expanding laboratory opportunities to the Internet. In 1998 professor Jesus del Alamo developed the Microelectronics WebLab, which allows students in one undergraduate and one graduate course, as well as students of the Singapore-MIT Alliance, to complete lab experiments remotely. The students carry out measurements online by controlling lab equipment through the Internet.

“I wanted my students to experience how transistors and other microelectronic devices actually work and compare that with the models and behavior I teach in class,” says del Alamo, who explains that prior to this, students had been unable to have a firsthand experience because the equipment is expensive, space is limited, and labs are difficult to manage. Now, however, more than 800 undergraduates and graduate students have had that experience through WebLab.

Next, del Alamo plans to enhance the experience by adding mechanisms for long-distance collaboration and for conducting simulations. Also, del Alamo says a new system architecture will make the lab easier to maintain.

Links Within

When it comes to attracting the best students, Guttag says his department is uniquely positioned to be competitive. “Electrical engineering and computer science are both very diverse fields today, and we’re one of the few departments anywhere that can bring critical mass to bear in a lot of aspects of those areas,” he says. At the same time, students have become more interested in studying both electrical engineering and computer science, instead of one or the other. Enrollment in Course VI-2, the track that combines these disciplines, has been climbing over the past few years. Meanwhile, the department’s graduate student enrollment is skyrocketing, with almost 3,000 applications for the doctoral program submitted this year.

With the move to the Stata Center this fall, the interaction between the two sides of the department will increase. “For the first time in our history, electrical engineering and computer science will basically be together geographically” Guttag says. “I think it will have a very big impact on how we teach and what research we do if we’re all close together physically.” Additionally, the Laboratory for Computer Science, the Artificial Intelligence Laboratory, and the Laboratory for Information and Decision Systems, which had been physically separated from the rest of the department, will soon be located just down the corridor from the academic department that supplies most of their researchers.

Victor Zue, director of the Laboratory for Computer Science, says the move into the Stata Center is “enormously important.” In terms of interacting with the rest of the campus, he says, “proximity makes all the difference. Moving us over to that part of the world will definitely have an impact towards enhanced collaboration with the electrical engineering side of the department.”

Plans are also in the works to merge the Laboratory for Computer Science and the Artificial Intelligence Lab, which already collaborate on a considerable amount of research, Zue says. The merger and subsequent renaming of the new, combined lab should happen midyear.

With increased collaboration both within the department and with other disciplines throughout the Institute, electrical engineering and computer science should continue to set the pace in research and teaching. On May 23 the department will celebrate its past as well as its current achievements at a centennial celebration. The event includes a day-long symposium in Kresge Auditorium that will highlight education and research initiatives. It will be followed by a reception and gala dinner at the Park Plaza Hotel.

Uh oh–you've read all five of your free articles for this month.

Insider Online Only

$19.95/yr US PRICE

Computing

From the latest smartphones to advances in quantum computing, the hardware behind today's digital age is rapidly changing.

You've read of free articles this month.