Issuu on Google+

SPRING 2013 Annual News from the MIT Department of Electrical Engineering and Computer Science

the MIT EECS

Connector

1

4

2

5

3

6

7


Annual News from the MIT Department of Electrical Engineering and Computer Science

the MIT EECS Connector Perspectives from the Department Head : 1 Department Snapshot : 3 SuperUROP Starts Strong : 7 Rising Stars in EECS : 11 Women’s Technology Program : 13 Centers: hubs for collaborative action : 14 bigdata@CSAIL : 14

Wireless@MIT : 15 Connection Science and Engineering : 17 MIT/MTL Center for Graphene Devices and 2D Systems : 18 Center for Excitonics : 19

Research Lab News : 21 CSAIL: From the Integrated Circuit to the Internet: Bridging Engineering and the Social Sciences, Constantinos Daskalakis : 21 CSAIL: Computer Aided Programming: Changing the way we code, Armando Solar-Lezama : 23 LIDS: When the whole is weaker than the sum of its parts: robustness and fragility in power grids, Mardavij Roozbehani, Munther Dahleh : 24 MTL: Nanofabrication, Karl K. Berggren, Vitor R. Manfrinato, Samuel M. Nicaise, Jae-Byum Chang : 26 RLE: Largest Ever Optical Phased Array, Michael Watts : 28

Anantha P. Chandrakasan Department Head Munther A. Dahleh Associate Department Head William T. Freeman Associate Department Head CONTACT the MIT EECS Connector Room 38-401 77 Massachusetts Avenue Cambridge, MA 02139 newsletter@eecs.mit.edu Editor: Patricia A. Sampson Design: Subbiah Design Printer: Artco, Inc.

www.eecs.mit.edu

Faculty News : 30 Awards, Fellowships, Chairs : 30, 33, 35

New Faculty : 38 L. Rafael Reif, MIT’s 17th president : 39

New Classes in EECS : 40 6.S02, a medical technology alternative to 6.02 is launched : 40

6.S193: Biological Circuit Engineering Laboratory : 42 6.S196 brings out the human side of technology : 44

Staff Features : 46 Claire Benoit, Agnes Chow, Janet Fischer and Myron (Fletch) Freeman Student Groups : 51 The MIT Formula SAE Team Goes Electric! : 51

6.270: The Course 6 autonomous robot competition – entirely run by students : 52 USAGE 2012–2013 : 54

Alumni Features : 55 Deborah Estrin PhD ’85 : 55

Perspectives from the Department Head

Drew Houston ’05, Arash Ferdowsi '08 : 57 Sal Khan SB, MEng ’98 : 60 Susie Wee SB ’90, SM ’91, PhD ’96 : 63

Donor Recognition : 66 Remember This? : inside back cover Front cover images: 1: Students in the Department of Electrical Engineering and Computer Science work on a lab project to build a light-tracking “pet robot” in 6.01. See Department Snapshot, page 3. 2: Read about the largest ever, optical phased array – work presented by Prof. Michael Watts in the Research Lab News, page 28. 3: The SuperUROP Starts Strong – read about the new Undergraduate Research Opportunities Program allowing in-depth high-caliber research and more on page 7. 4: Participants including top young female PhD graduates and postdocs gathered for the Rising Stars in EECS two-day workshop to present their research and network. Read more page 11. 5: Prof. Asuman Ozdaglar discusses connection science ideas with graduate students in her research group. See Department Snapshot, page 4. 6: Prof. Anantha Chandrakasan with collaborators used mammalian inner ear’s natural battery to power implantable devices. 7: EECS Professors William Freeman, Frédo Durand and John Guttag teamed to develop a system that amplifies video. See the EECS homepage www.eecs.mit.edu Feb. 28, 2013.

Welcome to the 2013 edition of the MIT EECS Connector. I am pleased to share news of ongoing academic and research programs as well as the implementation of the 2012 EECS Strategic Plan. Over the past academic year we launched several major initiatives that have already made a positive impact on our faculty, students, staff, and alumni. The EECS Department at MIT continues to lead internationally in education and research, establishing the basis for tomorrow’s breakthrough technologies. The new SuperUROP program, launched in September 2012, features a year-long advanced research experience during which undergraduate students (juniors and seniors) focus on a challenging research problem. The program provides mentorship and resources necessary to produce publication-worthy results and advanced software or hardware prototypes that could be commercially developed. In essence, SuperUROP is a jump-start on graduate school, a startup accelerator and an industry-training bootcamp, all rolled into one. With the generous financial support of industry and private donors, named undergraduate research and innovation scholars are engaged in projects that are commensurate with graduate level work. Students are also exposed to best research practices through the newly created class, 6.UAR: Preparation for Undergraduate Research. Several major inter-disciplinary centers have been created by EECS faculty over the past several years. The centers hosted by the EECS-affiliated research labs (CSAIL, LIDS, MTL and RLE) bring together faculty, students, staff, and sponsors from industry and government to address critical emerging problems in big data, wireless communications, connection science, materials and devices, synthetic biology, and health care. Several research centers are featured in this edition and more details are available on the lab websites. Our faculty continues to be recognized by major international awards. Several members of our faculty have been appointed to prestigious career development and senior faculty chairs. In order to recognize EECS faculty members for outstanding research contributions and international leadership in their fields, the Department has established the EECS Faculty Research Innovation Fellowship (FRIF) program. In 2013 we marked the second year of FRIF awards to support the research of senior faculty who do not hold endowed chairs. The newly established Steven and Renée Finn Innovation Fellowship provides tenured mid-career faculty in EECS with resources for up to three years to pursue new research and development paths. Another key initiative of the 2012 EECS Strategic Plan is Rising Stars in EECS, designed to bring together outstanding women PhD students close to graduation as well as women postdocs in electrical engineering and/or computer science. The first Rising Stars in EECS workshop, held November 2012, provided a two-day intensive networking experience for 36 attendees to share their research results

MIT EECS Connector — Spring 2013

1


Department Snapshot with each other and our faculty, and to learn about entering and building an academic career. Another goal is to increase their visibility with top departments in EE and CS. The workshop appears to have been a positive and energizing experience for all involved.

"the Department has created a flagship sophomore-level class, 6.S02: Introduction to EECS in Medicine."

In response to growing interest in bio-medical systems, the Department has created a flagship sophomore-level class, 6.S02: Introduction to EECS in Medicine, which is in its first offering this spring term. The class is designed to expose students to a wide spectrum of concepts, problems and hands-on laboratory experiences relevant to EECS in medicine. Our faculty members continue to engage in the pioneering MITx class offerings, including 6.002x (Circuits and Electronics) and 6.00x (Introduction to Computer Science and Programming). The Department is also offering many exciting new undergraduate courses, including Principles and Practice of Assistive Technology, Biological Circuit Engineering Laboratory (BioCel), Introduction to Inference, and Introduction to Machine Learning. We continue to engage our alumni in a number of programs. They are invited to provide cooperative teaching and mentoring through the Course 6 Mentoring Network, which is ideal for our large undergraduate software courses, such as 6.172, 6.005, and soon 6.813 and 6.170. The associated infrastructure, currently in development, will allow distant alumni to interact with students as reviewers of code. In this edition of the Connector we are very pleased to feature several alumni, including Dropbox co-founders Drew Houston (SB ’05) and Arash Ferdowsi ('08); Khan Academy founder Sal Kahn (SB ‘98, MEng ‘98); Susie Wee (SB’91, SM’91, PhD’96), Vice President and Chief Technology Officer of Networked Experiences at Cisco Systems; and Deborah Estrin (PhD ’85), Professor of Computer Science at Cornell Tech in New York City. It is my pleasure to share, through this edition of the Connector and through our website (www.eecs.mit.edu), the energy and excitement generated by our faculty, students and staff. We invite our alumni to be in touch through the Alumni News section of our website. I am also always delighted to hear directly from you. Best wishes, Anantha P. Chandrakasan Joseph F. and Nancy P. Keithley Professor of Electrical Engineering Department Head, MIT Electrical Engineering and Computer Science

EECS places renewed emphasis on interdisciplinary research, partnerships with alumni and industry, and experiential learning. Larry Hardesty, MIT News Office. November 16, 2012

Students in the Department of Electrical Engineering and Computer Science work on a lab project to build a light-tracking "pet robot" in 6.01. [Top left and center, photos by Dominick Reuter]. Undergraduate student Victor Pontis studies in a common area in the Stata Center (right). [photo by M. Scott Brauer. Both photos courtesy MIT News]

In the 1950s, when MIT researchers were helping to invent the discipline of computer science, they didn’t think of themselves as computer scientists; they thought of themselves as electrical engineers or physicists or mathematicians. Operating systems and programming languages were just tools they needed in order to maximize the productivity of the hugely complex new machines they were building. By 1975, however, computer science had developed enough autonomy that MIT’s Department of Electrical Engineering changed its name, becoming the Department of Electrical Engineering and Computer Science (EECS). Now, the Computer Science and Artificial Intelligence Laboratory (CSAIL) is the largest lab at MIT. EECS may now be in the midst of a similar expansion of its intellectual boundaries. According to department head Anantha Chandrakasan, the Joseph F. and Nancy P. Keithley Professor of Electrical Engineering, some of its most exciting research lies at the intersections of EECS and other disciplines. That includes research on “big data” — techniques for making sense of the massive amount of information unleashed by the Web, by biological, medical and physics research, and by the financial industry — as well as energy and biomedical research. “More than a third of our faculty are interested in the biomedical space,” Chandrakasan says.

2

www.eecs.mit.edu

At the same time, Chandrakasan says, the core EECS curriculum is more popular than ever. EECS has long drawn the largest undergraduate enrollment at MIT, since the days when it was just EE. But this year, Chandrakasan says, enrollment in the department’s two introductory courses — 6.01 and 6.02, in MIT’s course-numbering system — reached an all-time high. “Nearly half of MIT undergraduates take 6.01, regardless of major,” Chandrakasan says.

Data deluge With big data, the field is, in large part, reaping what it sowed. Exponential increases in computing power, and simpler tools for exploiting it, have led to an explosion of online data. But as rapidly as computers have improved, gene-sequencing machines have improved even more rapidly. Meanwhile, physics experiments at the Large Hadron Collider can generate petabytes of data every day. Andrew Lo, the Charles E. and Susan T. Harris Professor of Finance at the MIT Sloan School of Management, who has been on the MIT faculty since 1988, last year accepted a secondary appointment in EECS and became a primary investigator in CSAIL. Recently, Lo has used techniques borrowed from computer science to mine credit-bureau data

MIT EECS Connector — Spring 2013

3


Department Snapshot intracranial pressure from noninvasive-sensor data such as ultrasound scans and blood pressure measurements, rather than requiring physicians to drill holes in their patients’ skulls. CSAIL’s John Guttag, the Dugald C. Jackson Professor of Computer Science and Engineering, and RLE’s Collin Stultz, an associate professor of biomedical engineering, have mined electrocardiogram data to more accurately diagnose patients at risk for heart failure; RLE’s Elfar Adalsteinsson, an associate professor of electrical engineering and computer science and health sciences and technology, and Vivek Goyal, an associate professor of electrical engineering and computer science, developed algorithms that could reduce the duration of MRI scans from 45 to 15 minutes.

Left: Graduate student Kimon Drakopoulos (in green) presents his work on the LinkedIn social network to members of Asuman Ozdaglar's (in red) research group in a lab in the Connection Science and Engineering Center. [Both photos: M. Scott Brauer, courtesy MIT News] Right: Senior Adwoa Boakye works alongside other students on a lab for course 6.002, Circuits and Electronics, in the fifth floor student lab in Building 38. The project was the first group lab and focused on measuring output in metal-oxide-semiconductor field-effect transistors (MOSFETs) to see if observed results match theoretical predictions.

and data about the transactions conducted by customers of financial institutions to more accurately predict the risk of default or delinquency. Lo is one of the researchers at bigdata@CSAIL, a new initiative led by professor of computer science and engineering Sam Madden. Madden investigates techniques for searching databases more efficiently and for interpreting sensor data from networks of cars, among other things. Other project participants include professor of computer science and engineering Piotr Indyk, whose new algorithm for calculating the discrete Fourier transform — developed with professor of computer science and engineering Dina Katabi — has a broad range of applications in the big-data context, and associate professor of computer science and engineering Rob Miller, who finds ways to enlist human aid in the execution of large information-processing tasks. As people store more of their data online, however, it becomes more vulnerable to attack. Nickolai Zeldovich, an associate professor of software technology and another member of bigdata@CSAIL, has, together with Frans Kaashoek, the Charles A. Piper Professor of Computer Science and Engineering, researched ways to plug security holes in web applications; Zeldovich and Katabi introduced a new way to prevent the interception of wireless transmissions. And cryptography luminary Shafi Goldwasser, the RSA Professor of Computer Science and Engineering, a two-time winner of the Association for Computing Machinery’s Gödel Prize for theoretical computer science, and most recently (with Professor Silvio Micali) winner of the 2012 ACM Turing Award, has developed

4

www.eecs.mit.edu

algorithms that protect data stored in the cloud from particularly ingenious attacks. The Cryptography and Information Security Group at CSAIL contains celebrated luminaries. Silvio Micali, the Ford Professor of Engineering, shared the first-ever Gödel Prize with Goldwasser for the development of zero-knowledge proofs. The group’s other faculty members, adjunct professor of computer science and engineering Butler Lampson and Ron Rivest, the Andrew and Erna Viterbi Professor of Computer Science and Engineering, are recipients of the Turing Award, commonly referred to as the Nobel Prize of computer science. (In all, eleven researchers have won the Turing — Professors Shafi Goldwasser and Silvio Micali on March 13, 2013.)

The search for alternative sources of energy is squarely within the purview of classical electrical engineering: RLE researchers such as associate professor of electrical engineering Mark Baldo and Vladimir Bulovic´, the Fairborz Maseeh (1990) Professor of Emerging Technology, for instance, are developing flexible, transparent and even printable solar cells, while Bulovic´ and Jing Kong, the ITT Career Development Associate Professor of Electrical Engineering, showed that graphene — an atom-thick layer of carbon atoms — could offer a much more cost-effective way to provide electrodes for such devices. Tomás Palacios of the Microsystems Technology Laboratory, the Emanuel E. Landsman Career Development Associate Professor of Electronics, has shown that using gallium nitride in power converters that switch between alternating and direct current could cut mechanical devices’ power loss by 30 percent. Associate department head Munther Dahleh, professor of electrical engineering and computer science, on the other hand, is approaching the energy problem less directly. Among other things, Dahleh investigates how the type of control principles studied at the Laboratory for Information and Decision Systems can be brought to bear on management of the power grid.

Biology and energy

Undergrads as innovators

Other CSAIL researchers, such as associate professor of computer science Manolis Kellis and professor of computer science and engineering David Gifford, are developing novel algorithms for finding biologically informative patterns in mountains of genetic data. But another central area of convergence between computer science and medicine is the analysis and interpretation of signals from biomedical sensors.

The department’s undergraduate curriculum, too, has an increasingly interdisciplinary flavor. Its introductory courses, 6.01 and 6.02, concentrate on robotics and communications, respectively, canvassing a wide range of topics — from control theory and algorithms to signal processing and circuit design. Chandrakasan says the department is planning to offer a third introductory course, which will concentrate on biomedical applications of EECS principles.

For instance, CSAIL’s Polina Golland, an associate professor of computer science and engineering, finds correlations between anomalies in brain scans and neurological disorders. Similarly, George Verghese of the Research Laboratory of Electronics (RLE), the Henry Warren Ellis Professor of Electrical Engineering, has developed algorithms that could infer changes in

The creation of a third introductory course is an element of Chandrakasan’s strategic plan for the department, which EECS leaders began developing soon after he became department head in 2011. The plan’s hallmark educational initiative, however, is the so-called “SuperUROP” program, which builds

Top: EECS postdoc Puneet Srivastava (right) works with Mark Mondol, facility manager at the MIT Electron Beam Lithography lab, at MIT in Cambridge, Mass. Srivastava is learning how to use the tool.Bottom: Junior Sylvia Zakarian (center) works alongside other students on a lab for course 6.002, Circuits and Electronics. [Both photos: M. Scott Brauer, courtesy MIT News Office]

on MIT’s much-emulated Undergraduate Research Opportunities Program (UROP). Founded in 1969, UROP offers funding and academic credit to undergraduates who do original research in MIT labs. While the majority of UROP projects last only a semester, SuperUROP projects last a full year, and students are required to take a yearlong course in which a series of outside speakers discuss both research topics and entrepreneurship. Each student receives a stipend for the year, and each faculty supervisor gets additional funding in his or her lab budget. The yearlong student commitment, and the additional research funding, makes sponsoring a SuperUROP student much more attractive to faculty, Chandrakasan says; this greater faculty involvement, in turn, enriches the undergraduates’ experience. The SuperUROP program launched this fall — with funding from both private donors and a roster of 14 corporate sponsors — and 86 EECS undergraduates enrolled. At the beginning of the year, the program’s website posted a detailed list of more than 100 research projects that faculty were willing to supervise; the corporate sponsors posted a second list. But several students opted instead to create their own projects and find faculty to sponsor them. Indeed, Chandrakasan says, one of the program’s aims is to provide an outlet for the entrepreneurship that seems to be second nature for many MIT students. MIT EECS Connector — Spring 2013

5


Department Snapshot

SuperUROP Starts Strong

By Lauren J. Clark

Similarly, Miller has recruited MIT alumni to help with code reviews in a course he teaches, 6.005 (Elements of Software Construction). There, however, the mechanism is much different. Where the volunteers in Leiserson and Amarasinghe’s course meet with students in pairs, those in Miller’s course examine small chunks of code online whenever they have time. Indeed, cultivating an “alumni teaching network” is another aspect of EECS’s strategic plan, Chandrakasan says. “We want to engage the broad network of our alumni, but also experts out there in industry,” he says. The department is also in the process of building a new Engineering Design Studio, led by professor of mechanical and electrical engineering and computer science Steven Leeb. Chandrakasan describes it as “a hands-on advanced-prototyping lab where students can prototype both mechanical and electronic systems, including biomedical devices. We’ll have state-of-the-art wireless components and devices from electronic-systems manufacturers. We’ll have tutorials from industry people through videoconferencing, teaching students how to build and use systems.” The studio would be available for lab assignments in undergraduate courses and to undergraduates pursuing independent projects, Chandrakasan says. Top: Senior Robert Johnson works on final project ideas on a blackboard in the 38-600 student lab at MIT. Lower left: Harvard-MIT Division of Health Sciences and Technology graduate student Gabrielle Merchant works with Teaching Assistant David Jenicek (right) alongside other students on a lab for course 6.002, Circuits and Electronics. Lower right Students listen as Professor Dennis Freeman speaks to the 6.UAR students about the EECS MEng program [All Photos: M. Scott Brauer, courtesy MIT News]

Meet the real world The SuperUROP initiative illustrates another principle that, Chandrakasan argues, has been crucial to the success of EECS at MIT: corporate partnership. Both CSAIL and MTL have existing programs for corporate sponsors — respectively, the Industry Affiliates Program and the Microsystems Industrial Group. Both programs provide companies with opportunities to sponsor research, to initiate joint projects, and to keep abreast of the labs’ research. In turn, the programs help match MIT students with prospective employers. Recently, Chandrakasan points out, some CSAIL professors have begun to engage with industry in novel ways. Three years ago, professors of computer science and engineering Charles Leiserson and Saman Amarasinghe, who co-teach 6.172 (Performance Engineering of Software Systems), initiated a program that recruits senior programmers from the Boston area to review the code written by students in the class. In the program’s first few years, Leiserson has said, he and Amarasinghe had to turn volunteers away.

6

www.eecs.mit.edu

Equalizing opportunities Even as EECS takes steps to enhance the educational experience for MIT students, it’s also assumed a leading role in edX, an ambitious project to make the educational resources of MIT, Harvard University, the University of California at Berkeley, and the University of Texas system — and, over time, other universities — available to students worldwide. The first course offered through edX was an EECS class, 6.002x (Circuits and Electronics), and professor of computer science and engineering Anant Agarwal, the CSAIL head who stepped down to become president of edX, led a group of EECS researchers in developing the technological platform that undergirds the whole system. Since the introduction of 6.002x, Chandrakasan says, on-campus enrollment in 6.002 has gone up by 40 percent. In addition to its educational initiatives, the department’s strategic plan also aims to increase the number of women faculty in electrical engineering and computer science. The centerpiece of that project is “Rising Stars in EECS,” an annual workshop, created by Chandrakasan and Golland, that brings talented women from the nation’s top EECS graduate programs to the MIT campus for two days of symposia and talks on research and job-search skills. The first such workshop was held earlier this month. n

Left: Professor Anantha Chandrakasan talks with students at the SuperUROP class 6.UAR, fall 2012. Middle: Analog Devices Chairman Ray Stata ’57, SM ’58 talks with the 6.UAR students about startups and building a company. Right: 6.UAR students and guests listen to a panel discussion, fall 2012. [Photos by Bethany Versoy]

In September 2012, EECS and the Undergraduate Research Opportunities Program (UROP) launched SuperUROP, which has the potential to transform undergraduate education. It gives EECS juniors and seniors the chance to make greater contributions to MIT’s world-renowned research than they would under a typical UROP. “This is an amazing experience as an undergrad,” says Luis Voloch, who is investigating how computer networks conceal and reveal sources of information. “SuperUROP is a special arrangement, because there’s a high level of expectation from the students. It creates a setting in which professors and students take a project very seriously and for a long term. I think it’s a good preview of what graduate school can be like.” Voloch’s research is supported by Draper Laboratory. When Anantha P. Chandrakasan became department head in July 2011, he spearheaded the expansion of research opportunities for undergraduates. The idea for SuperUROP evolved out of his 2012 strategic plan, and he invited the Undergraduate Student Advisory Group (USAGE) to play a lead role in developing a program that excited fellow students. An impressive 86 juniors and seniors have participated in SuperUROP’s pilot year, and interest among both students and faculty has been so great that the program will likely expand to other departments. Launched in 1969, when undergraduates were rarely found in the laboratory, UROP evolved into a widely copied MIT program in which more than 80 percent of the Institute’s undergraduates participate. Typically, they spend a semester getting their feet wet in the laboratory, experiencing what it’s like to work side-by-side with senior researchers.

But many students extend their UROP projects — often for a year or more — indicating that there is a demand for greater exposure to the rewards and complexities of scientific investigation. “Many students desire a more in-depth research experience — one that culminates in results that could be published in a journal or top conference, or advanced prototypes that could be commercially developed,” says Chandrakasan, who is also the Joseph F. and Nancy P. Keithley Professor of Electrical Engineering. SuperUROP is, in essence, a jump-start on graduate school, a startup accelerator, and an industry-training bootcamp, all rolled into one. In October, EECS held a reception to celebrate the launch of SuperUROP. Students in the inaugural class, EECS faculty, individual and industry sponsors and MIT administrators who helped implement the program gathered in the Stata Center R&D dining area for the festivities. Chandrakasan welcomed all at the gathering and acknowledged those who helped make SuperUROP a reality. He noted that the program would not be possible without significant financial support. Fourteen companies, along with individual donors, generously support SuperUROP through the Research and Innovation Scholars Program (RISP). RISP is a prestigious named-scholars program that underwrites the participating students and provides some discretionary funding for the host research group. The companies and donors provide not only mentoring, but also project suggestions and research directions. Some of the key players in implementing SuperUROP gave remarks at the reception. They included Julie Norman, MIT 

MIT EECS Connector — Spring 2013

7


SuperUROP Starts Strong “The exciting thing about MIT is the number of revolutionary ideas that develop here. I think it’s important to encourage [the students] — for them to see that people from the real world who invest are putting in the time to see what they’re working on and that [research is] not just academic,” Grinnell says. Many of the students are working on research aligned with the interests of SuperUROP’s industrial sponsors. These companies aim to solve complicated, open-ended problems ranging from achieving ultra-low power computing and nanoelectronics systems to modeling the massive amounts of data generated by social networks. Rui Jin, a senior, is taking a mobile-phone charging technology developed by Texas Instruments and specializing it for medical devices. Representatives of the company, which supports Jin’s research through RISP (Research and Innovation Scholars

Along with Colin and Erika Angle and several anonymous donors, Dinarte R. Morais and Paul Rosenblum, Jr., both 1986 EECS alumni, provide financial support for SuperUROP. Industrial sponsors are Analog Devices, Basis Technology, Denso, Draper Laboratory, eBay Inc., Facebook, Foxconn, Google, Intel, MediaTek, Qualcomm, Quanta Computer, Texas Instruments and VMware. “As an industrial sponsor, Analog Devices will look for opportunities to collaborate with students and faculty on research topics of continual interest and provide insights into the relevance of research to real-world applications,” says Stata. “Analog Devices is excited about exploring new possibilities to strengthen our relationship with MIT students and faculty through the SuperUROP program.”

8

www.eecs.mit.edu

really own and really take control of” Program), attended the poster session and were “pretty enthusiastic,” he says." Jin explains why his research requires the time, training and technical facilities that SuperUROP provides.

Many SuperUROP students work on projects proposed by MIT faculty, who are eager to bring enterprising undergraduates into their laboratories. Faculty submitted 100 different research topics for the program’s inaugural year, including a microelectromechanical tactile display for the blind, an integrated language for Web application coding that would fix security vulnerabilities and coding mistakes, and a method of predicting a hospital patient’s blood pressure based on cloud-scale machine learning.

High-caliber research — and much more In addition to a research experience that lasts a year or more, SuperUROP participants take a two-semester class, “Preparation for Undergraduate Research,” that covers topics ranging from industry best practices to presentation skills to ethics in engineering.

Notably, in their roles as UROP supervisors and academic advisors, faculty can have a major impact on a student’s decision to attend graduate school as well as the graduate school selection process. Among seniors choosing to further their education, 68 percent indicated that faculty had provided assistance in their search for a graduate program.

They also get access to MIT’s sophisticated nanofabrication facilities (through the Microsystems Technology Laboratories) — a privilege typically reserved for graduate students. Upon completion of the program, they receive a certificate in advanced undergraduate research with a designated focus area, such as artificial intelligence, computer systems or nanotechnology.

Lyne Tchapmi Petse is deciding whether to pursue graduate school or an industry position. She has been working with EECS Professor Charles Soldini on a smart phone-compatible heart-monitoring system.

SuperUROP students receive a significant stipend of $3,000 per semester for 10 hours per week of work. Their faculty supervisors receive $4,000 to support the student for the entire academic year. The students have the opportunity to interact with industry mentors and venture capitalists. In December, Fairhaven Capital founder and EECS alumnus Rick Grinnell ’92, SM ’93 attended SuperUROP’s first poster session to meet the program’s participants and offer them advice.

independent project that I can

“With my project, I need to first build a prototype, but that’s not the end of it. I need to test the prototype and make improvements. But the ultimate goal is to actually fabricate silicon and design a chip that will support all the features that are in my prototype. All of that work combined will take far more than one semester—more than even one year,” Jin says, adding that he will expand his SuperUROP project into his graduate thesis.

Left: Speakers at the Oct. 18, 2012 reception for the SuperUROP included, from left, Counterpoint Health Sciences CEO, Dr. Erika N, Angle ’04, Analog Devices Chairman Ray Stata ’57, SM ’58, EECS Department Head Anantha Chandrakasan, Julie Norman, MIT senior associate dean for undergraduate education and director of undergraduate advising and academic programming, Carine Abi Akar, ’12, member of USAGE in 2011-12, and Dennis Freeman, EECS professor and undergraduate officer. [Photo: Patricia Sampson] SuperUROP student Annie Holladay was featured in the MIT News Office on Feb. 25, 2013, for her contributions to work in the Learning and Intelligent Systems Group at MIT’s Computer Science and Artificial Intelligence Laboratory showing how household robots could use lateral thinking to compensate for their physical shortcomings. [Photo: Allegra Boverman, courtesy MIT News]

senior associate dean for undergraduate education and director of undergraduate advising and academic programming; Analog Devices Chairman Ray Stata ‘57, SM ‘58; Counterpoint Health Sciences CEO Dr. Erika N. Angle ‘04, who with her husband, iRobot Corporation CEO and chairman of the board, Colin A. Angle ‘89, SM ‘91 supports SuperUROP; Carine Abi Akar, ‘12, member of USAGE in 2011-2012; and Dennis Freeman, EECS professor and undergraduate officer.

“SuperUROP offers me an

Top: SuperUROP students Avanti Shrikumar and Jennifer Wang pick out a T shirt at the SuperUROP reception in the MIT Stata R&D Dining area. Bottom: Prof. Arvind, left, talks with SuperUROP student Stephan Boyer at the Dec. 6, 2012 SuperUROP poster session. [Photos: Patricia Sampson/EECS]

Worn just behind the ear, the device transmits vital signs in real time via Bluetooth radio signals to a smart phone app for analysis and display. “You have Bluetooth on your smart phone, so you can basically have a device that you can wear at home and that can help you monitor your heart,” says Petse, whose work is supported by Analog Devices.

MIT EECS Connector — Spring 2013

9


SuperUROP Starts Strong

Rising Stars in EECS

By Lauren J. Clark

Another student, Sebastian Leon, is among the very first researchers to examine the user dynamics of edX, which offers free online courses from MIT, Harvard and other universities. A member of EECS Professor Una-May O’Reilly’s research group, Leon explains that edX has about 100,000 users whose interactions with the platform generate a “gigantic mass of raw data. The idea is to create a predictive model of student behavior based on all of this data. “SuperUROP offers me an independent project that I can really own and really take control of,” Leon adds. MIT Chancellor and EECS faculty member Eric Grimson is inspired by such commitment to high-caliber work. “The level of research being conducted is remarkable, and the articulate manner in which students talk about their research and the excitement they communicate are impressive,” said the Bernard M. Gordon Professor of Medical Engineering during SuperUROP’s December poster session. In addition to working on industry- and faculty-led research projects, students may choose their own topics. In fact, choosing a research topic and pitching it to a VC or a government sponsor such as DARPA is part of the syllabus of “Preparation for Undergraduate Research.” However they get involved in their research projects, undergraduates can gain a lot from SuperUROP, says Stan Reiss of Matrix Partners, another EECS alumnus and venture capitalist who attended the poster session. “They can get to some real results, and that’s the kind of experience you really need,” he says. “Even if [the students] have no intention of commercializing their work, this is a great program. It’s an opportunity to do something relevant to the real world—to whatever they’re going to end up doing when they graduate.” Whether they go on to graduate school, a startup or an industry career, SuperUROP gives MIT students a valuable head start on generating the revolutionary ideas of the future. n

The Rising Stars in EECS participants gathered with MIT EECS faculty and visiting faculty and administrators from the University of California at Berkeley and the University of Rochester to present their research and network with each other for two days in November 2012. Here they took a moment to pose on the fourth level patio of the Stata Center. [Photo: Patricia Sampson/EECS]

On Nov. 1 and 2, nearly three dozen of the world’s top young female electrical engineers and computer scientists gathered at MIT to experience something rare: they outnumbered the men in the room. The MIT Electrical Engineering and Computer Science Department invited the women to its inaugural “Rising Stars in EECS” workshop. Attendees came from MIT, Stanford University, the University of California at Berkeley, Cornell University, Carnegie Mellon University, the Max Planck Institute in Munich, Ecole Polytechnique Federale de Lausanne in Switzerland and other research institutions to network with one another and with faculty from MIT and elsewhere. On the cusp of entering the workforce, these PhD candidates and postdocs came for guidance on launching careers as professors and to raise their visibility in the field of electrical engineering and computer science. “You see so few women [in EECS], it’s nice to see them all together,” said Lydia Chilton, a PhD candidate at the University of Washington who studies crowdsourcing and other aspects of human-computer interaction. “As I’ve gotten older I really value the female colleagues that I have. I feel like I interact with them more naturally.”

Top: SuperUROP students Jelimo Maswan, left, and Lisa Liu discuss their research at the Dec. 6, 2012 poster session. Middle: SuperUROP student Jeff Chan discusses his research with a fellow SuperUROP student. Bottom: SuperUROP student Arvind Thiagarajan discusses his research with SuperUROP Industrial Interface Coordinator Ted Equi at the Dec. 6, poster session. [Photos by Patricia Sampson/EECS]

10

www.eecs.mit.edu

“I’m at the end of my PhD, and I thought this was an amazing opportunity to meet other women interested in education and to get some advice,” said Floraine Berthouzoz, who works in computer graphics and computer vision at UC Berkeley and is the only woman in a research group of about 20 people.

While women’s representation in most science and engineering fields has increased substantially over the past 25 years, their participation in EECS has been halting. A recent National Science Foundation survey shows that women make up just 22 percent of PhDs in computer science. Moving up the ranks of academia, the numbers become even more stark: women comprise a mere 10 percent of tenure-track faculty in the top electrical engineering academic departments, according to the National Academies. “When you talk to search committees at different universities, one of the complaints is that there aren’t many applications from women,” said Polina Golland, an MIT associate professor in EECS who organized Rising Stars with the head of her department, Anantha P. Chandrakasan, the Keithley Professor of Electrical Engineering. They plan to hold the workshop annually. EECS departments at MIT and peer institutions are keen to add more women to academia’s pipeline by identifying potential candidates and encouraging them to apply for faculty positions, said Golland. She believes that the workshop was a significant step in the right direction. It afforded a valuable opportunity, she said, for elite researchers to meet peers in the broad field of EECS—outside the confines of their respective sub-disciplines—and to be inspired by established women faculty. Kristen Dorsey, a PhD candidate studying microelectromechanical systems at Carnegie Mellon, said: “Before attending [the workshop], I thought, ‘Well, I’d like to be a MIT EECS Connector — Spring 2013

11


Rising Stars in EECS

Left: EECS Professor Dina Katabi introduces several Rising Stars presenters. Rising Stars participants also presented their work in several poster sessions in the Bill and Melinda Gates Tower of the Stata Center (right top and lower middle photos). They met for the conference in the Star Conference room of the Alexander W. Dreyfoos, Jr. Tower of the Stata Center. [Photos: Patricia Sampson/EECS]

faculty member, but I don’t have what it takes. I don’t have the publication record, I don’t know what I’m doing.’” However, being invited to MIT and meeting women who have successful careers in academia gave her a measure of confidence, she said.

women faculty, including National Medal of Science honoree Mildred Dresselhaus and Turing Award recipient Barbara Liskov. They also got a primer on how the promotion process works from a panel of faculty representing Harvard, MIT, Boston University and the University of Rochester.

Dorsey and other attendees, including Jenna Wiens of MIT, added that meeting talented researchers from the wide variety of fields that EECS encompasses exposed them to new possibilities for collaboration and professional support.

The issue of work-life balance arose at the panel on promotions and at occasional moments throughout the workshop. The panelists noted that it has become common policy at universities to adjust the “tenure clock” for women who have to take time off to care for infants, and that male faculty can and do take advantage of family leave policies.

“It’s a great networking opportunity to meet other women who are interested in pursuing careers in academia whom I could forge relationships with early on—and then hopefully tap into that network later,” said Wiens, whose research focuses on machine learning and data mining. That was precisely Chandrakasan’s vision when he began planning Rising Stars with Golland as part of his department’s 2012 strategic plan. He was inspired by the success that the Department of Aeronautics and Astronautics has had with a similar annual workshop for women. In his welcoming remarks, he told the invitees, “As you start thinking about applying for faculty positions, I hope you can use each other as a resource. This is a group I hope will stay together.” The workshop’s attendees shared their research during formal presentations and poster sessions on topics ranging from improving online video streaming rates to tamper-proofing circuits to modeling the risk of infection among hospital patients. They networked over breakfast with MIT senior 12

www.eecs.mit.edu

During the Q&A session with junior faculty, Dana Weinstein, MIT’s Steve and Renée Finn CD Assistant Professor in EECS, pointed out that while being in a high-profile academic position keeps her very busy, “there is lot of flexibility in your schedule. You can have a life outside of work.” Looking ahead to a career as a faculty member, Chilton of the University of Washington was enthusiastic about other ways in which academia offers flexibility. “I like that it doesn’t preclude me from doing other things, such as a startup. And I really like being on the cutting edge. I feel like, even though I’m making things that aren’t immediately practical, they will be in five years.” It is Chandrakasan’s and Golland’s bet that a cohort of passionate and talented women like Chilton will advance the field of EECS—not only through groundbreaking research, but by paving the way for the next, substantially larger, generation of female engineers and scientists. n

EECS Women’s Technology Program

Students and staff of the 2012 EECS Women's Technology Program

“Not only have I learned so much about electrical engineering, computer science, and discrete math, I have also gained a broader perspective on the cuttingedge discoveries and breakthroughs that are happening at MIT and developed my ability to think critically to solve problems.”

Since 2002, the EECS department has reached into the high school pipeline to attract young women to engineering and computer science with the Women’s Technology Program (WTP). This four-week academic and residential experience allows female high school students to explore EECS through hands-on classes, labs, and team-based projects in the summer after 11th grade. The goal of WTP is to spark girls’ interest in future study of engineering and computer science. This introductory yet rigorous program is taught by female MIT graduate students and is designed for high school girls who excel at math and science, have the requisite background to be engineers and attend MIT, but who are not yet headed in that direction. The EECS department provides essential resources to support this important and successful outreach mission. WTP-EECS has tracked the college career pathways of its alumnae over 10 years to measure the program’s impact: 65% have majored in a field of engineering or computer science; another 20% have majored in a science or mathematics. 43% have matriculated at MIT, with 73% selecting majors in the School of Engineering and 29% majoring in Course 6. To learn more about WTP visit: http://wtp.mit.edu n

MIT EECS Connector — Spring 2013

13


Centers: hubs for collaborative action Dealing with Big Data The bigdata@CSAIL initiative was launched at MIT on May 31, 2012 to focus on managing the huge amount of information generated today by modern enterprises, websites, and networked sensors. Bigdata@CSAIL includes a number of industry partners, including AIG, Alior Bank, EMC, Intel, Huawei, Microsoft, Samsung, SAP and Thomson Reuters. Professor Sam Madden of the MIT Department of Electrical Engineering and Computer Science (EECS) heads bigdata@ CSAIL, and Elizabeth Bruce acts as its executive director.

When bigdata@CSAIL launched in spring 2012, Massachusetts Governor Deval Patrick announced the formation of the Massachusetts Big Data Initiative. This initiative will sponsor a grant-matching program, an internship program, and a project to investigate how big-data technologies can improve government. Another large area to explore includes security in relation to information policy. What is data privacy? As bigdata@CSAIL grows, it is hoped that outside collaborators— such as major research hospitals, for example — will benefit from the data

As Sam Madden pointed out at the May 31st launch event, “existing tools for analyzing data are outdated and rooted in computer systems and technologies developed in the 1970s.” This is the core mission of the 20 EECS colleagues and six MIT CSAIL researchers and staff in bigdata@CSAIL: to develop a new generation of technologies to store, manage, analyze, share, and understand today’s huge quantities of data.

In terms of defining big data, Executive Director Elizabeth Bruce says, “From the technology side, you can define big data with the three “V’s”: volume, velocity — collecting data real-time; and variety.” She explains that the latter is considered one of the

14

www.eecs.mit.edu

experts in more highly facilitated research. Elizabeth Bruce points to several projects that kicked off this past fall that are making use of one of the world’s largest cancer patient databases located at Massachusetts General Hospital. Locating this major research initiative at MIT CSAIL is also a way to ensure that education in this new field is a natural by-product. Encouraging interested students and building data scientists will build the new field. Elizabeth Bruce explains: “With the growth of big data, we also want to demonstrate the importance of open access to data. Some of the questions that are going to come up in the future are probably going to be questions of ownership. Transparency is important. We would like to show the good of big data — how data can be used to really improve the quality of our lives.” n

Wireless@MIT is created to address immediate and long-term wireless needs

The multidisciplinary approach of bigdata@CSAIL will bring together world leaders in parallel architecture, massive-scale data processing, algorithms, machine learning, visualization, and interfaces to identify and address four research themes: first, scalable platforms — computing infrastructures that can expand indefinitely to accommodate increasing data loads; second, the data-analysis tools that will run on those platforms; third, scalable techniques for handling security and privacy; and fourth, new techniques for visualizing and understanding data. Work in these four areas is being coupled with application experts in areas such as finance (Professor Andrew Lo), medicine (Professor John Guttag), science (Professor Stonebraker), education (through a relationship with the MITx initiative), and transportation (Professor Hari Balakrishnan and Professor Madden). To date, over twenty MIT faculty-headed big data projects are underway including: Energy-Efficient Algorithms (Erik Demaine); Uncovering Clinically Relevant Medical Knowledge (John Guttag); Execution Migration Machine (EM2) (Srini Devadas); Maintaining Coherent Memory in Dynamic Distributed Systems (Nancy Lynch); Privacy-preserving Methods for Sharing Financial Risk Exposures (Andrew Lo); BLINKDB (Sam Madden); Smart Transportation Cartel (Hari Balakrishnan, Sam Madden); and Energy: Hydrocarbon Exploration (Piotr Indyk, Tommi Jaakkola, William Freeman and Tomaso Poggio).

history) of the data and visualizing the data. Elizabeth Bruce points out that the data will need to be accessible to not just a handful of experts — but, in a large corporation, for example, to multiple levels from corporate leadership on down to allow the whole organization to benefit from the analytics. This will be true for multiple areas including government.

Last October, the Computer Science and Artificial Intelligence Lab (CSAIL) launched a new interdisciplinary center dedicated to developing the next generation of wireless networks and mobile devices. The MIT Center for Wireless Networks and Mobile Computing dubbed "Wireless@MIT," headed by EECS faculty members Hari Balakrishnan and Dina Katabi, offers a unique approach to advancing wireless communications by bringing together researchers and companies from across the wireless ecosystem.

Top: The May 2012 launch of the new initiative bigdata@CSAIL brought crowds to hear speakers including Governor Duval Patrick, MIT President Susan Hockfield, Intel Chief Technology Officer Justin Rattner, CSAIL Director Daniela Rus and EECS Professor and leader of bigdata@CSAIL Sam Madden. Bottom: Professor Sam Madden, bigdata@CSAIL leader is flanked by MIT President Susan Hockfield (left) and Intel CTO Justin Rattner. [Photos: Jason Dorfmann/CSAIL]

most difficult challenges. Increasingly, there are many kinds of data – images (photos and videos that will account for over 85% of total Internet traffic in the next few years, for example); and Twitter and other social network feeds, all of which need to be integrated, structured and then analyzed. In addition, the challenges of developing the technologies to store the data must be addressed—once it has been determined what to store, verifying and assuring the reliability (the

Industry partners in the new center include Amazon.com, Cisco, Google, Intel, MediaTek, Microsoft, STMicroelectronics, and Telefonica. The Wireless@MIT collaboration is aimed at influencing and impacting standards and products while offering a medium for companies, academics and government representatives to discuss the future of the wireless industry. Wireless@MIT co-director Dina Katabi, professor of electrical engineering and computer science, notes about the structure of the center: “If you look, there are many centers in other universities and institutions but they are focused on a specific area. Some centers only work on circuits, others work only on applications. The unique feature of this center is that it brings the experts in radios, circuits, communications, networks, and mobile apps, and makes them work together on innovative technologies." She adds, “ A lesson we have learned in wireless is that you cannot have breakthrough innovations by working on one aspect of the problem alone.”

Today the wireless industry is composed of entities—such as application vendors and content providers, network operators, equipment manufacturers and radio chipset developers—operating more or less independently of each other. Wireless@ MIT will bring these groups closer together with researchers to enable a more coordinated, holistic approach to mobile system design. It is not too soon for this development. “There are already over five billion mobile phones in the world today; add to this all the tablets, laptops, medical devices and wireless sensors, and the numbers are staggering,” says Hari Balakrishnan, the Fujitsu Professor in Electrical Engineering and Computer Science, and co-director of the center. “The goal of our center is to push the frontiers of wireless research to their full potential, and to ensure that the industry that grows up around these new devices is able to work in innovative and productive ways.” MIT EECS Connector — Spring 2013

15


Centers: hubs for collaborative action Wireless@MIT’s aggressive goals start with the spectrum crisis – the exhaustion of radio spectrum caused by the explosive popularity of wireless systems. Prof. Katabi notes: “By getting more from the spectrum than we already have, we can get 10 times higher data speeds for our wireless networks, and we can do it all without asking for additional spectrum.” She adds, “And, we do have technologies that can get you potentially toward this magnitude over the next five years.”

Connection Science and Engineering: a virtual center brings together network initiatives from across MIT holistic framework for the study of decentralized networks that are comprised of both human and technological elements.”

The second goal of Wireless@MIT is energy and energy efficiency. “Transmission energy is a humongous consumer of energy,” notes Prof. Katabi. The center develops hardware and software designs for energy-efficient mobile systems. These include low-power handsets, energy-scavenging sensors, and wireless medical devices – work that is carried out by Prof. Anantha Chandrakasan, the Joseph F. and Nancy P. Keithley Professor of Electrical Engineering and Arvind, the Johnson Professor of Computer Science and Engineering at MIT.

The Connection Science center will focus on fundamental research and applications within four broad areas: modeling social flows, understanding implications of networked interactions on economic and financial outcomes, developing a theory of network computation, and design of social architectures. The scope of social flows will include developing a systematic framework for modeling and understanding the nature and spread of information and influence in dynamically evolving and complex social networks.

The third goal of the center includes the development of a new paradigm for wireless video delivery. “We all want video on our wireless devices. And, this is one example where you can see the importance of having people work together,” says Prof. Katabi. “When you are trying to deliver that video to the user, it's not just the video, it's the channel and how you record the signal on the channel and the video application.” By looking at the whole system and coordinating the various layers required, Prof. Katabi notes, today’s uneven delivery will gain in performance, reliability and efficiency. The fourth goal of the center is delivering new means for secure website service across commercial and other Internet transactions. Projects underway include Hari Balakrishnan’s MOSH: Mobile Shell as a robust and responsive replacement for SSH particularly over WI-FI, cellular, and long-distance links. Another application known as Tamper-Evident Pairing (TEP) is being developed by Dina Katabi and EECS Associate Professor Nickolai Zeldovich. TEP is the first wireless pairing protocol that works in-band, with no pre-shared keys and protects again MITM (man-in-the-middle) attacks.

Top: Graduate student Ezzeldin Hamed (in white, seated) shows software radios to, from left, Professor Hari Balakrishnan, graduate student Shuo Deng, graduate student Anirudh Sivaraman and Professor Dina Katabi in the Wireless Center. [Photo: M. Scott Brauer, courtesy MIT News Office]. Middle: Wireless@ MIT demos in late May, 2012. On March 6, Julius Genachowski, Chairman of the United States Federal Communications Commission (FCC), answered questions about wireless spectrum - including spectrum sharing, spectrum access and allocation, and the impact of the spectrum crunch on the wireless industry - with Professors Hari Balakrishnan and Dina Katabi, co-directors of the MIT Center for Wireless Networks and Mobile Computing (Wireless@MIT). [Bottom two photos: Jason Dorfman/CSAIL]

16

www.eecs.mit.edu

Where will life with wireless take us in the future? Dina Katabi describes what is coming sooner than we think. “Cell phones are just one part of the revolution,” she says. “Whether they will completely morph to glasses or to whole environment; it’s like being able to have a wireless interface without carrying any device. Wireless signals reflected off our skin may be used to capture our gestures. Such technologies can change how we interact with devices in our environment and will enable many new interesting applications.” n

The center also aims to study the implications of networked interactions in shaping economic behavior and financial outcomes. This will include investigation of how social networks shape economic behavior, consumer choice, insurance and collective action. In addition, understanding how ideas, failures and risks spread and cascade through networks and how to leverage the power of networks to improve resource allocation and technological innovation will be addressed. In 2011, more than 20 senior MIT faculty members from all five schools came to discuss a new virtual center for connection science. That so many chose to attend this 8:30 am meeting on one week’s notice was considered a good portent for the new initiative.

Any modeling exercise commensurate with the size of today’s networks will inevitably run into computational challenges. An important goal of the center is to design scalable local algorithms for network computation and develop tractable algorithms for detection and inference in large networks.

This meeting marked the start of a new virtual center that is projected to lay the foundation for a new discipline as well—to be known as Connection Science. Engaging existing networkbased initiatives from across the Institute, Connection Science will focus on core theory to develop experimentally-based mathematical models that are powerful enough to predict network outcomes. Among the issues to be addressed, the initiative will also aim to build metrics to measure changes in network structure as well as construct models for multi-scale interactions among connected networks of different sizes and types. “Our lives have been transformed by networks that combine people and computers in new ways,” notes Connection Science center’s co-director Asu Ozdaglar, the Steven and Renée Finn Innovation Fellow and professor of electrical engineering and computer science in the MIT Department of Electrical Engineering and Computer Science. She adds: “The new center ‘Connection Science and Engineering’ at MIT brings together a large number of researchers working on different aspects of this increasingly connected landscape to develop a systematic and MIT EECS Connector — Spring 2013

17


Centers: hubs for collaborative action Social architectures entail design of new applications and architectures that enable more efficient and flexible interplay between technological and social platforms. Developing such architectures will facilitate and empower the user collective human-computer intelligence and large-scale human assisted computation. The new center also has an educational mission that includes development of new multi-disciplinary core courses to be available Institute-wide. Courses in the field will not only introduce fundamental tools for analysis and design of networked systems but also provide the bridge for PhD candidates to launch their research careers in this area including postdoctoral appointments.

The Connection Science and Engineering center held an interdisciplinary workshop titled “Information and Decisions in Social Networks” in November 2012. The workshop was widely attended with more than 200 participants from around the world. It featured three plenary talks by leading experts in the field on topics such as use of mobile phone communication networks for understanding network structure, compatibility networks in kidney exchange, and competitive contagion models on networks. The workshop also included a panel discussion moderated by Connection Science and Engineering Center co-founder and Media Lab Prof. Sandy Pentland, in which four leading thinkers from the communication, mobile, banking and information industries discussed the challenges and opportunities in network research. n

MIT/MTL Center for Graphene Devices and 2D Systems

new applications in electronics, chemistry and material science. The Center is also exploring two-dimensional materials beyond graphene, including insulators such as hexagonal boron nitride (h-BN) and semiconductors, like molybdenum disulfide (MoS2). "The unique structure and properties of two-dimensional materials have the potential to impact numerous industries," says Tomás Palacios, the Emanuel E. Landsman Career Development Associate Professor of Electronics at MIT's Department of Electrical Engineering and Computer Science, and first director of the new center. "The MIT/MTL Center for Graphene Devices and 2D Systems is an important driving force in exploring the numerous applications for these materials and in creating a vision for the future of graphene-enabled systems." The Center’s focus is highly interdisciplinary and it coordinates projects on new semi-transparent graphene electrodes for future solar cells and displays, MoS2 integrated circuits for

transparent electronics, graphene and BN membranes for water purification, 2D materials for a new generation of solid state lighting, to name only a few of the projects that are being pursued within the Center. This center benefits from very close collaboration with industrial partners. According to Michael Strano, Professor in the Department of Chemical Engineering and co-director of the center: "This academic-industrial partnership is essential to the advancement of both fundamental graphene science, and of emerging technological applications. One of the main goals of the Center is to create an environment that fosters this collaboration." The Center coordinates the work of the more than 15 MIT research groups in 6 different Departments, and leverages several existing collaborative efforts in 2D material science and engineering that currently exist on campus, including a Multidisciplinary University Research Initiative grant (MURI) from the Office of Naval Research, as well as a regular Boston-Area CarbOn Nanoscience (BACON) Meeting. For more information, please visit: www-mtl.mit.edu/wpmu/graphene/ n

Getting the Best out of Nature’s Physics Excitons have a lot of researchers very excited When a chlorophyll molecule in the leaf of a plant absorbs a photon of sunlight, the solar energy is converted into an excited state of the molecule known as an exciton. The exciton then transports the energy between molecules in the leaf and ultimately mediates the conversion of sunlight into electrical energy.

Although electronics has been playing a key role in our society for the last 60 years, its presence is still far from being ubiquitous. The great majority of the objects around us do not have any electronics in them. In addition, the very high cost of starting new semiconductor companies is severely limiting the potential growth of the electronics industry. The MIT/MTL Center for Graphene Devices and 2D Systems (MIT Graphene Center) was started in the second half of 2011 as an interdepartmental center and part of the Microsystems Technology Laboratories (MTL) with the mission of bringing together MIT researchers and industrial partners to develop a new generation of materials and devices that can potentially transform electronics.

18

www.eecs.mit.edu

Novel materials are key to developing the next generation of electronics and the MIT Graphene Center is focusing on the amazing properties of graphene and other two-dimensional materials. Graphene, a form of pure carbon arranged in a hexagonal lattice just one atom thick, has generated great excitement among researchers worldwide. Not only is graphene the thinnest material known, it is also the strongest. It has unique transport properties that allow unprecedented electron and hole mobilities. And being extremely thin, graphene is mechanically flexible, optically transparent and can be easily integrated with other materials. Until recently, most studies of this amazing material have focused on its basic physical properties. Work at the MIT Graphene Center is focused on how these amazing properties can be exploited in

Excitons are crucial to the development of low cost solar energy and energy efficient lighting. They are packets of energy confined within a material and the crucial intermediate for energy transduction in all kinds of low-cost electronic materials. Along with molecular systems like photosynthesis, they also dominate the behavior of synthetic nano-materials like polymers and inorganic quantum dots. Consequently, excitons control solar energy conversion in low-cost solar cells, and also light emission in organic and quantum-dot based LEDs. See the video “Excited About Excitons” – winner of the Life at the Frontier of Energy Research video contest. (techtv.mit.edu/videos/11732-excited-about-excitonics)

Excitonics Center is founded Founded in 2008, the Center for Excitonics is an Energy Frontier Research Center funded by the U.S. Department of Energy, Office of Science and Office of Basic Energy Sciences. Its mission is to develop the science and technology of excitons, to reveal the fundamental characteristics of these crucial quasi-particles, and enable new solar cells and lighting technologies. The Center’s work is divided into four working groups, each containing three to five faculty devoted to key scientific problems confronting the development of more efficient solar cells and state lighting. The Coherence and Disorder group’s work aims to understand and control coherence in excitonic antennas MIT EECS Connector — Spring 2013

19


Centers: hubs for collaborative action

Research Lab News From the Integrated Circuit to the Internet: Bridging Engineering and the Social Sciences by Constantinos Daskalakis, Associate Professor, CSAIL The integrated circuit is a good metaphor for what we do in Engineering. Its characteristics: high complexity, limited resources (e.g., space and energy), centralized design, and cooperative components — its gates, switches, capacitors, etc., interact as programmed to bring about a computation. Throughout Engineering we study systems with these characteristics (computers, cars, airplanes, factories, etc.), and have developed rich theories with which we can make predictions about the performance of a design to be able to adjust and optimize it. The Internet, however, does not fit the above paradigm. Firstly, its design is not centralized. While its basic protocols, such as the Border Gateway Protocol (BGP), were centrally designed and deployed throughout the Internet, its topology, for example, was not. Internet Service Providers (ISPs) are free to design their own network and connections to other ISPs, and each of us is free to connect or disconnect our own computer from the Internet. Moreover, the various entities that make up the Internet have free will, and will operate strategically to optimize their own objectives.

Transmission electron micrograph of 2 nm features using the STEM as the exposure tool.

and exciton polaritons (coherent combinations of excitons and photons). The Semiconductor Nanocrystals group works towards understanding exciton dynamics in semiconductor nanocrystals using multiexciton spectroscopy and photonic interrogation of single quantum dots in the visible and infrared. The Solar Antennas group seeks to use excitonics to collect, concentrate, and wavelength-convert sunlight for single junction solar cells. New states of matter and energy are created when excitons are coupled with photons – also called ‘exciton polaritons’. Exploring the new classes of energy conversion devices forms the basis for the Hybrid Excitonics. The work of any one group with the Excitonics Center spills over to others. For example, EECS graduate student Vitor Manfrinato as a member of EECS Prof. Karl Berggren’s Quantum Nanostructures and Nanofabrication Group at MIT has collaborated with researchers at the Brookhaven National Laboratory in New York (a partner with the Excitonics Center) to use the large scanning transmission electron microscope (STEM) there to create circuits for excitons. The scale of these circuits is measured in quantum dots (QDs) — roughly eight atoms wide. And, the idea is to ‘write’ these structures in microseconds. Using electron beam lithography (EBL), Vitor’s research process uses beams of electrons to define patterns on a smaller scale than even optical lithography can. He is changing the materials and techniques to find ways to improve production efficiently at the scale of individual atoms. Ultimately, Vitor hopes to position the QDs exactly as needed so that engineering the transfer of energy between the dots will permit new solar devices. Read more about this work in the Research Lab Features, page 28. n

20

www.eecs.mit.edu

(from left to right) Eric Stach, Dong Su, Lihua Zhang, Vitor Manfrinato at Brookhaven National Laboratory with STEM microscope.

For instance, would an ISP always use BGP as prescribed? Recent observations prove that this is not always the case: recall the recent global two-hour outage in YouTube access, which resulted from a mere BGP table update — a result of censorship — in Pakistan [2]. Ultimately, the various systems developed on the Internet provide new environments for social (think facebook, Linkedin, Twitter) and market (think eBay, online advertising) activity, adding socio-economic characteristics to it. Daskalakis researches how to incorporate socio-economic considerations into Engineering to address the challenges arising on the Internet. A possible approach to do this would be to import tools from Game Theory and Economics straight into Engineering. When a system’s behavior is unpredictable due to strategic interactions among the system’s users or administrators, we may appeal to the appropriate game-theoretic tool to figure out what is going to happen. This way, we can engage in our usual iteration of system design, behavior prediction, design modification, new behavior prediction, and so forth, until the system displays the desired behavior. So what tools may we want to use for this purpose? Game Theory offers several. The crown-jewel, going back to the work of John Nash in 1950 for which he won the Nobel prize in Economics, is the Nash equilibrium. This is a state of

Professor Costis Daskalakis speaking at TEDx Athens in 2011.

operation in which every interacting party is doing the best for itself given the actions of the other parties. If a system finds itself in such a state, there is no incentive for unilateral changes of behavior taking the system to different states. The postulate, often employed in Game Theory, is that systems eventually reach equilibria, and we may study them at equilibrium: the Nash equilibrium, or one of its many refinements or generalizations. But here is where the plot thickens. How long may the transient phenomena last before the system reaches equilibrium? Is there a reason to expect transient phenomena to be short? ...so that we may ignore them and focus on equilibrium states for the purposes of predicting system behavior? MIT EECS Connector — Spring 2013

21


Research Lab News Some thought reveals that these are computational questions, as strategic interaction is computation of sorts. Using tools from computational complexity theory, Daskalakis established that Nash equilibria are, computationally speaking, unreachable [1, 3]. The result is existential; it shows that systems exist that will not reach Nash equilibrium in centuries of computation time. And this casts doubt on the universal applicability of equilibrium theory for the purposes of predicting behavior – since, for any notion of practical time, systems may operate outside of equilibrium.

proposing a novel algorithmic framework for auction design [4]. The interaction of Computer Science, Game Theory, and Economics has already been fruitful, providing algorithmic insights to fundamental problems in Game Theory and Economics, and equipping Computer Science with tools for understanding strategic behavior such as that observed on the Internet. How this interaction will shape the future of these fields, and how it will affect online activity is to be explored. n

Are there structural properties of a system that would make equilibria reachable? Or should we dismiss equilibrium theory, and replace it with some other theory? Without answering these questions we have our hands tied in creating a robust Internet, since we cannot predict how individuals interacting in a system will behave and what effects their strategic behavior will have in the system’s overall behavior. Daskalakis and his students are investigating these questions focusing on ways to navigate system design away from conditions that enable computational complexity to kick in. They are prototyping this approach in the design of auctions, which play a central role in today’s online activity.

[1] Constantinos Daskalakis, Paul W. Goldberg and Christos H. Papadimitriou. The complexity of computing a Nash equilibrium. Communications of the ACM, 52(2):89–97, 2009.

We are all familiar with sponsored search results, appearing on the right hand side of a web search with google, bing etc. These results are advertisements, and what advertisements are displayed and in what order is decided by an auction, taking place in the milliseconds between clicking “search” and receiving the results. Similarly, auctions are used to decide what advertisements are shown in webpage banners. And there are direct uses of auctions in online markets such as eBay. When designing an auction, the auctioneer allocates resources to bidders and receives payments from them looking to optimize some objective: fairness, social happiness, or revenue. The challenge is that bidders will try to manipulate the outcome of the auction through their bids, shifting it away from optimality. Thus, the auction must be cleverly designed to cancel the effects of manipulation. For example, how does one sell a single item to optimize revenue? Is the auction employed by eBay optimal or are there better auctions? In 1981, economist Roger Myerson showed that eBay is essentially optimal. His work was very influential and won him the 2007 Nobel prize in Economics. But it left unanswered the question of how to design optimal auctions for allocating multiple items — the problem facing sponsored search and display advertising. Unresolved since the 1980s, this problem has been one of the most important in the field of auctions. Tying together techniques from Economics and Combinatorial Optimization, Daskalakis and his students recently provided a solution, generalizing Myerson’s result to multiple items, and

22

www.eecs.mit.edu

References

[2] Pakistan blocks YouTube website, BBC News Online, February 24, 2008. http://news.bbc.co.uk/2/hi/south_asia/7261727.stm [3] What computer science can teach economics, MIT News, November 8, 2009. web.mit.edu/newsoffice/2009/game-theory.html [4] Computer science tackles 30-year-old economics problem, MIT News, June 2012. web.mit.edu/newsoffice/2012/comp-sci-econ-0625.html

Computer Aided Programming: Changing the way we code by Armando Solar-Lezama, NBX Career Development Assistant Professor, CSAIL The last ten years have seen the coming of age of a range of technologies to check for the absence of important classes of errors, but checking for errors is only the first step down the road to better, more reliable software. The same technology that can help us discover bugs and verify the correctness of existing software can enable a new class of programming systems that will make programming easier and may ultimately lead to better software that is both more reliable and easier to produce. In the Computer Aided Programming group, we have been developing a language called Sketch as a vehicle to explore these ideas. The defining feature of Sketch is the ability to leave holes in programs: in place of complex expressions, the programmer can provide the set of building blocks from which an expression can be constructed, and the system will explore the set of all possible expressions that can be produced from the building blocks until it finds one that matches the specification. For example, Figure 1 shows an example of a partition function that partitions a range [0,N) among P processors. Given a processor id pid, the function computes the beginning and end of the range for that processor. Instead of writing the function in full, the programmer specifies the building blocks for the different expressions, and the synthesizer automatically derives the correct expressions. In order to describe to the system what the function is supposed to do, the user provides a test harness that checks that the resulting partitions are balanced, and that they cover the entire range.                        

Figure 1

Sketch can find a correct code fragment among an astronomically large space of possible expressions in a matter of seconds. The key is to perform the search symbolically;

v next prev

a

next prev

next prev

a

x

next prev

next prev

b

b

next prev

next prev

Figure 2

instead of searching through possible expressions one by one, Sketch produces a system of equations that represents the correctness criteria and uses a combination of symbolic manipulation and Boolean satisfiability solving in order to find a solution to these equations. The use of symbolic search has been used very effectively in the past to find bugs in programs. By extending this idea to search program fragments, we can make programming easier by discovering many of the details that make programming difficult. The reach of this technology extends beyond simply allowing us to leave holes in programs; the ability to search very large spaces of code fragments is an enabler for new programming methodologies that are more intuitive and less error prone. For example, graduate student Rishabh Singh has developed a system on top of Sketch called the Storyboard Programming Toolkit (SPT) that can take a graphical description of a data-structure manipulation—similar to the balls-and-arrows diagrams programmer's commonly draw in a whiteboard—and synthesize low-level code from it. In another project, graduate student Alvin Cheung has developed a system called QBS that can synthesize SQL code taking an imperative code fragment as a specification. This allows the system to automatically shift functionality from the application to the database, a task that traditionally required significant programmer effort. The technology behind Sketch can also make it easier to teach programming. Rishabh Singh has been leveraging the ability to search large spaces of programs to find corrections to programming assignments submitted by students. In the case of small programming errors, the technique can precisely pinpoint the error and suggest a correction that will make the error disappear. In the case of conceptual errors, for example due to a misunderstanding of the problem statement, the technique can identify the conceptual error and expose it to the student. In short, the ability to infer correct code fragments that satisfy certain correctness conditions has the potential to transform not just the way we program, but also the way we teach programming in the future. n MIT EECS Connector — Spring 2013

23


Research Lab News When the whole is weaker than the sum of its parts: robustness and fragility in power grids by Mardavij Roozbehani, Principal Research Scientist and Munther Dahleh, Professor Laboratory for Information and Decision Systems It is hard to notice all the transformative changes that are taking place in the electric power grids around the world when you are working in your office and the lights are steadily on. And, unless you were on the sample path of one of the recent blackouts, the statistics of power outages in the United States might surprise you. In fact, data from NERC (North America Electric Reliability Corporation) and EIA (US Energy Information Administration) show that both frequency and size of the power outages in the United States have been steadily on the rise.

system as the system components respond to exogenous disturbances. For example, the tripping of protective relays isolates faults, but also creates a buildup of load on the rest of the system sometimes in a way that can lead to a cascade of other faults. Therefore, while such systems may perform well under normal operating conditions or be very robust to most common disturbance scenarios, they can exhibit fragility in response to certain large or small disturbances. Such fragility in power systems and more broadly in interconnected systems is characterized by systemic risk or endogenous risk. At an abstract level, fragility of a feedback control system can be explained by some of the results in the classical control literature, which establish tradeoffs between performance and robustness. In view of these tradeoffs, fragility can be interpreted as follows: increasing robustness with respect to certain types of disturbances, inherently increases sensitivity with respect to uncertainties and disturbances that are not included in the model. Such concepts are however, less well understood, less developed, and less formalized for interconnected systems. Our research is motivated by both the theory gap for characterization of such tradeoffs in networked systems, and by the need to address a pressing and increasingly more important problem: systemic risk and cascaded failures in power systems.

Figure 1: Statistics of blackouts in the US

Although some recent large blackouts have been caused by natural disasters, e.g., hurricane Sandy, it is rare that a large and systemic failure results from a single unbearable disturbance that brings down an entire system. Usually, such large failures are the result of an increased level of risk or reduced level of robustness, which brings the system closer to a state of failure or makes it more vulnerable to a sequence of contingencies, certain patterns of disturbances, or in some cases, a single moderate shock. The contingencies can be the result of a variety of events, such as volatile weather events, local component failures, intermittencies in renewable generation, etc. But the cascades that follow such events are often reinforced or amplified by mechanisms that are put in place to improve efficiency under normal conditions, or increase robustness to withstand other types of disturbance. In other words, fragility builds up in the 24

www.eecs.mit.edu

Part of our research in this area has focused on analysis of the feedback loops and the interactions between the market layer (where pricing and economic decisions take place) and the physical layer (where power flows from generators to loads) of power systems.

agents have complete knowledge of how their decisions affect the market price, and are fully rational in strategizing their decisions to minimize their expected cost. By characterizing the statistics of the stationary aggregate output process across a spectrum of networks from fully cooperative to fully non-cooperative, we showed that a tradeoff exists between efficiency (aggregate system cost) and risk (tail probability of aggregate output). While the non-cooperative network leads to an efficiency loss — widely known as the "price of anarchy" — the stationary distribution of the corresponding aggregate output process has a smaller tail, whereas, the cooperative network achieves higher efficiency at the cost of a higher probability of output spikes. Furthermore, the cooperative network has a smaller output variance, which can be interpreted as higher robustness; but it also has a higher probability of large output spikes, which can be interpreted as higher fragility. This particular fragility emerges when a large accumulated backlog in the system coincides with lack of flexibility to either absorb the backlog, or schedule it for the future. Intuitively, the cooperative scheme allows for shifting flexible loads more aggressively, thereby increasing the probability of a large backlog in the system, which eventually leads to a large spike. The noncooperative scheme however, is more conservative in shifting flexible loads, which, as a result, is less likely to lead to a large backlog (Figures 2 and 3). This discussion motivates questions such as “What is a right measure of systemic risk?”, “How do we estimate the risk of a large-scale system like the grid?”, “How does network connectivity determine fault propagation”, and “How can we develop design principles to balance the trade-offs and achieve efficiency under normal operation, while enabling early detection, containment of risk, and graceful degradation upon approaching a state of failure?” These issues constitute the core of our research. n

In our earlier research, we showed that volatility increases in the system when price-taking consumers actively respond to wholesale electricity market prices due to the uncertainty in their response to price signals, and uneven sensitivity to prices induced by threshold policies. Our recent work has revealed more subtle sources of fragility in power systems. Together with EECS graduate student Qingqing Huang, we developed developed abstract models that show how fragility and endogenous risk can be inherent to the architecture of the system, and arise from the dynamics of the system even under the most ideal assumptions of fully rational agents with perfect information. Together with EECS graduate student Qingqing Huang, we developed a model of a dynamic oligopolistic energy market in which, a set of distributed agents with market power dynamically update their output (consumption or production) decisions. In this model, the

Figure 3: Sample paths of the aggregate output process. At a smaller time scale (top), the cooperative load scheduling can smooth out the aggregate output process. However, at a larger time scale (bottom), there are more output spikes produced endogenously by the cooperative load-scheduling scheme.

Figure 2: Conceptual diagram of efficiency-risk frontier.

MIT EECS Connector — Spring 2013

25


Research Lab News Nanofabrication by Karl K. Berggren, Associate Professor; Vitor R. Manfrinato, Samuel M. Nicaise, Jae-Byum Chang, Microsystems Technology Laboratories, Research Laboratory of Electronics Nanolithography, the principle means by which nanometer-length-scale patterns are placed on a surface, is a cornerstone of the modern microelectronics industry and is integral to the future of nanotechnology. But no one can yet reliably pattern features with adequate resolution, speed, uniformity, and, most of all, sufficiently low cost, to make the nanotechnology of tomorrow a reality today. To solve this problem, several nanolithography techniques are available, each with strengths and weaknesses. Electron-beam lithography (EBL — pattern generation using nanometer-sized electron beams) has the highest resolution, but is not particularly well-suited to high-volume production; instead it is applied principally to device prototyping and lithographic-mask manufacturing. Ion-beam lithography (pattern generation using nanometer-sized ion beams) can achieve similar resolution to EBL, but with increased exposure efficiency. But ion-beam lithography tools are less well developed than EBL tools, and suffer from many of the same problems. Self-assembly of nanomaterials — the thermodynamically driven arrangement of patterns on a surface without the use of lithography — is extremely low-cost and fast due to its parallel nature. However, self-assembly struggles in producing ordered patterns over long length scales. This

disorder has recently been overcome by combining self-assembly with EBL, in an approach known as templated or directed self-assembly (TSA or DSA). The Berggren group has collaborated with Prof. Caroline Ross, in the Dept. of Materials Science and Engineering, and used DSA to quickly generate high-resolution patterns with good long-range order. The Berggren group also performs fundamental research on electron- and ion-beam lithography to increase the resolution, yield, and quality of nanofabrication with these methods, and to apply this knowledge to various electronic and photonic applications. In particular, the group investigates using nanostructured arrays fabricated by EBL as templates for TSA, and for controlled placement of nanometer-sized light emitters, such as colloidal quantum dots. In a recent attempt to achieve the ultimate resolution limit of EBL, Berggren’s group performed EBL with the smallest electron-beam spot size available (in collaboration with Brookhaven National Lab). In lithography, the exposure tool is used to affect chemical change in a thin resist layer on a substrate. The resist is then developed to produce a topographic pattern. The team used an aberration-corrected scanning transmission electron microscope with 0.15-nanometer-diameter electron-beam spot size as the exposure tool.

The team also used a high-resolution resist and did all the subsequent metrology with a transmission electron microscope. The results were images of 2-nm-wide features, roughly 10 atoms wide, the smallest structures ever achieved in a conventional electron-beam lithography material (Figure 1A and 1B).

local area. Control of long-range and uniform nanopatterns were achieved with templates of particular spacings and angles (Figure 2). These long-range ordered two- or threedimensional patterns could in principle be used for fabricating a wide array of electronic and photonic devices.

The group is also looking for alternative charged-particle species that could be potentially higher resolution and faster than EBL. Berggren’s group, in collaboration with Carl Zeiss NTS, recently demonstrated that it is possible to write nanostructures with neon ions. Neon-ion-beam lithography has resolution comparable to EBL (10 nanometer line width) but an exposure efficiency 1000× greater than EBL at comparable electron landing energies (Figure 1C). Thus, neon could be a viable agent for production of nanopatterns for masks for lithography, or even perhaps for direct writing of nanostructures on a device wafer.

One example of an emerging scientific field that particularly benefits from nanolithography is nano-optics. Nano-optics consists of capturing, processing, and emitting light at the nanometer scale—essentially it is engineering antennas for visible light. One promising nanometer-sized light source is a colloidal quantum dot (QD). These QDs are low cost and solution-processable materials, with fine synthetic control of their electronic and optical properties at the sub-10 nanometer length-scale. However, precise placement of QDs is needed in order to design nano-optical systems at the single-dot level. Berggren’s group, in collaboration with the group of Moungi Bawendi, the Lester Wolfe Professor of Chemistry at MIT, developed a technique to control the placement of 5-nanometer-diameter QDs at a desired position with EBL (Figure 3). This placement technique resulted in an average of three QDs at any given desired position. Furthermore, these QDs could be placed in close proximity to one another, with a minimum separation of 12 nm. Photoluminescence (specific light emission by previously absorbed light) measurements show that the placed QD clusters are optically active after the fabrication process. This optimized top-down lithographic process is a step towards the integration of individual QDs in new electronic and optical systems. n

The Berggren group’s efforts in template self-assembly have focused on using block copolymers (BCP). Polymers are comprised of long chains of repeating, similar, molecules. BCPs are interesting because two different molecular sections (blocks) of the polymer repel each other. Self-assembly then results in small domains that repeat with separations on the 10s-of-nanometer scale. The chemical differences between the small domains make them great candidates for self-assembling complex patterns. Unfortunately, BCPs do not naturally self-assemble into the long-range-ordered patterns that would be most useful to nanopatterning. To control the self-assembly of BCP, so as to fabricate useful patterns, we used EBL to create topographic post arrays which attract only one of the blocks. The choice of which block would be attracted to the post, as well as the arrangement of posts in the array, determined the orientation, spacing, morphology, and layering of the BCP. Complex nanopatterns could thus be achieved with templates that directed the self-assembly in a

Figure 1. (A) and (B) Top-down images obtained using a transmission electron microscope of the features (dark regions) fabricated using 200 keV electrons in a scanning transmission electron microscope. (C) Top-down image obtained using a scanning electron microscope of the features (horizontal bright lines) fabricated using neon-ion beam lithography. [Photos: courtesy of the Berggren group]

26

www.eecs.mit.edu

Figure 2. Top-down images obtained using a scanning electron microscope. In (A) the block copolymer (alternating light grey and dark regions) self-assembled to form a nanopattern with two predominant directions. Templating posts (white dots) directed the BCP into this nanopattern. (B) is similar to (A), though the pattern formed by block copolymer is a mesh formed from a top and bottom layer (light and dark grey respectively). Again, the templating posts (white dots) directed the self-assembly of the BCP in both of the layers. Notice that the BCP nanopattern outside of the templated region is not a mesh pattern. [Photos: courtesy of the Berggren group]

Figure 3. Top: EBL design used to fabricate a hole array for placing sub-15-nm clusters of colloidal quantum dots (QDs). Center: Photoluminescence (light emission generated by previously absorbed light) image of the positioned QDs. Bottom: Top-down image obtained using a scanning electron microscope of placed QD clusters (each QD is a bright circular spot). [Photos: courtesy of the Berggren group]

MIT EECS Connector — Spring 2013

27


Research Lab News Largest-Ever, Optical Phased Array By Michael Watts, KDD Career Development Associate Professor in Communications and Technology Research Laboratory of Electronics A phased array consists of several typically identical antennas, with each antenna emitting signals characterized by a specific amplitude and phase that interfere to form a desired radiation pattern in the far field. Typically phased arrays are tunable and can be used to rapidly redirect electromagnetic or acoustic radiation in a desired direction. It has been more than a century since the first radio-frequency (RF) phased arrays were invented. Today, RF phased arrays have seen widespread adoption in applications ranging from radar to the wireless communications networks that empower our smartphones enabling bandwidth to be locally provisioned. And, acoustic phased arrays have been used for underwater and ultrasound imaging. Similarly, the ability to generate high-resolution and arbitrary radiation patterns at optical frequencies would have a profound impact on a number of important applications. One of the principle near-term applications is for scanning LADAR (Laser Distance And Ranging) beams in autonomous vehicles for accident avoidance. However, the range of applications extends well beyond this realm. From optical communications to biomedical imaging and ultimately 3-D holographic displays, optical phased arrays are expected to have wide-ranging impact. The short optical wavelength nevertheless imposes stringent requirements on chip fabrication, as even nano-scale fluctuations in device dimensions significantly affect the optical emission from the nanoantennas. Consequently, all demonstrations of optical phased arrays so far have been restricted to 1-D, or small-scale (16 antenna) 2-D arrays. Recently, Prof. Watts’ research group, the Photonic Microsystems Group, has developed an advanced silicon photonics platform within a state-of-the-art 300mm CMOS fabrication facility, using standard CMOS electronics fabrication techniques. Coupling these advanced processing techniques with some unique designs, Prof. Watts has recently demonstrated phased arrays at optical frequencies composed of 4,096 nanoantennas integrated on a small silicon chip. The results are described in the January 10, 2013 issue of Nature (Vol. 493, No. 7431, pp. 195-pp.199) and consist of a 2-D nanophotonic phased array, in which 64 x 64 optical nanoantennas are integrated on a silicon chip. With a dense footprint of 576μm x 576μm, the 4,096 nanoantennas are precisely balanced in power and aligned in phase to generate a designed and sophisticated radiation pattern—in this case the MIT-logo. A conceptual diagram of this nanophotonic phased array is depicted alongside a micrograph of the fabricated structure in

28

www.eecs.mit.edu

Fig. 1 (below). Light is successively tapped off of a common silicon input waveguide to couple to the individual nanoantenna elements. Due to wiring limitations, this large phased array was hard-encoded by adjusting the length of the S-bend structures in each cell. Near-vertical emission was then produced by a strong grating written into the silicon waveguide, forming the nanoantenna.

Figure 2. (left) Near-field and (right) far-field patterns of the 64×64 nanophotonic phased array. The MIT-logo was formed in the far field through interference, as designed.

Figure 1. (left) Schematic of the passive large-scale nanophotonic phased array. (right) Micrograph of the nanophotonic phased array. [All Figures: Jie Sun, postdoctoral associate, RLE]

Despite the lack of a phase adjust knob, the far-field pattern accurately reproduced the MIT-logo (Fig. 2) (left column, below), highlighting the robust design and fabrication process. However, active phased arrays with the ability to generate dynamic radiation patterns in the far field were also demonstrated in a somewhat smaller, 8 x 8 element array, depicted in Fig. 3 (next page). Active tuning of the array was achieved through the use of heater elements placed directly in the silicon waveguides and driven by a copper interconnect. As seen in Fig. 3 (next page), beams can be steered in the vertical and horizontal directions, representing the first nanophotonic phased array steerable in both dimensions. The robust design of the array, coupled with state-of-the-art complementary metal-oxide-semiconductor technology, allows large-scale optical phased arrays to be implemented on compact and inexpensive nanophotonic chips. In turn, the novel device architecture and fabrication process extend the functionalities of phased arrays, opening up possibilities for large-scale deployment in a wide variety of applications, including 3-D holography, biomedical sciences, LADAR, and communication systems. n

Figure 3. (top) Schematic of the active 8×8 nanophotonic phased array. (bottom) By applying different voltages on the phased array, the optical beam can be steered in both the vertical and horizontal directions.

MIT EECS Connector — Spring 2013

29


Faculty News : Awards

30

Scott Aaronson

Hal Abelson

Anant Agarwal

Arvind

Bonnie Berger

Rodney Brooks

2012 NSF Alan T. Waterman Award

ACM 2012 Karl V. Karlstom Outstanding Educator Award

Elected to National Academy of Engineering

Elected to American Academy of Arts and Sciences 2012 IEEE Computer Society’s Harry H. Goode Award

2012 Fellow of the International Society for Computational Biology (ISCB)

2013 IEEE Fellow Selected by EE Times as Visionary

Anantha P. Chandrakasan

David Clark

Costis Daskalakis

Jesús del Alamo

Erik Demaine

Jack Dennis

IEEE 2013 Donald O. Pederson Award [pictured] ISSCC Top Contributor Awards [at the 60th Anniversary]

ACM Test of Time Award (with Karen Sollins, Principal Research Scientist,CSAIL)

2012 Microsoft Research Faculty Fellowship

2012 Intel Outstanding Research Award SRC Technical Excellence Award IEEE Electron Devices Society Education Award

2013 European Assoc. for Theoretical Computer Science (EATCS) Presburger Award for young scientists

2013 IEEE von Neumann Award

Mildred S. Dresselhaus

James Fujimoto

Qing Hu

Frans Kaashoek

Barbara H. Liskov

Nancy Lynch

2012 Kavli Prize in Nanoscience

2012 Antonio Champalimaud Vision Award

IEEE Photonics Society’s 2012 William Streifer Scientific Achievement Award

Elected to American Academy of Arts and Sciences

Elected to National Academy of Sciences Charter Fellow of National Academy of Inventors

2012-13 ACM Athena Lecturer

www.eecs.mit.edu

MIT EECS Connector — Spring 2013

31


Faculty News : Innovation Fellowships

Faculty News : Awards

Faculty Research and Innovation Fellowships for 2012

Timothy K. Lu

Wojciech Matusikk

Rob Miller

2012 Presidential Early Career Award for Scientists and Engineers

2012 DARPA Young Faculty Award

2013 MacVicar Faculty Fellow

Frédo Durand

David Perreault

Yury Polyanskiy

Ronald Rivest

2013 IEEE Fellow

NSF Career Award

National Cyber Security Hall of Fame

Piotr Indyk

The Electrical Engineering and Computer Science Department at MIT has announced that the recipients of the Faculty Research and Innovation Fellowship (FRIF) for 2012 are Frédo Durand, Piotr Indyk and Pablo Parrilo. In its second year, the FRIF is given to recognize senior EECS faculty members for outstanding research contributions and international leadership in their fields. Winners of the FRIF receive three years of gift funding. [Photo: The 2012 Faculty Research Innovation Fellows are (left to right) Frédo Durand, Piotr Indyk and Pablo Parrilo.] Frédo Durand works both on synthetic image generation and computational photography. His research interests span mostly aspects of picture generation and creation, with emphasis on mathematical analysis, signal processing, and inspiration from perceptual sciences. Frédo has been recognized for his work including an inaugural Eurographics Young Researcher Award in 2004, an NSF CAREER award in 2005, and an inaugural Microsoft Research New Faculty Fellowship in 2005. Frédo is a member of CSAIL.

32

Nir Shavit

Dana Weinstein

Victor Zue

2012 Edsger W. Dijkstra Prize in Distributed Computing

Intel Early Career Faculty Honor Program

2013 IEEE Flannagan Award 2013 Okawa Prize

www.eecs.mit.edu

Pablo A. Parrilo Pablo A. Parrilo works on optimization methods for engineering applications, control and identification of uncertain complex systems, robustness analysis and synthesis, and the development and application of computational tools based on convex optimization and algorithmic algebra to practically relevant engineering problems. Pablo has received several distinctions, including a Finmeccanica Career Development Chair, the Donald P. Eckman Award of the American Automatic Control Council, the SIAM Activity Group on Control and Systems Theory (SIAG/CST) Prize, and the IEEE Antonio Ruberti Young Researcher Prize. He is currently on the Editorial Board of the MOS/SIAM Book Series in Optimization. Pablo is an Associate Director of LIDS. Inaugural FRIF recipients in 2011 included Vladimir Bulovic´, Tommi Jaakkola, Dina Katabi and Muriel Médard. Read more about them in the 2012 EECS newsletter the Connector at: http://www.eecs.mit.edu/docs/newsletter/connector2012.pdf n

Piotr Indyk works on high-dimensional computational geometry, sketching and streaming algorithms, sparse recovery and compressive sensing. He is a recipient of the NSF CAREER Award, Sloan Fellowship, Packard Fellowship and Technology Review TR10. He is an Associate Editor for the IEEE Transactions on Signal Processing and SIAM Journal on Computing. Piotr is a member of CSAIL.

MIT EECS Connector — Spring 2013

33


Faculty News : Innovation Fellowship

Faculty News : Chairs Vladimir currently serves as Director of the Microsystems Technology Laboratories and is playing a critical role in defining the future of MIT’s nanofabrication capabilities. n

Asu Ozdaglar is the Inaugural Steven and Renée Finn Innovation Fellow 2012 Professor Asu Ozdaglar has been named the inaugural Steven and Renée Finn Innovation Fellow. Made possible by a gift from MIT Electrical Engineering and Computer Science alumnus Steven Finn '68, SM '69, EE '70, ScD '75 and his wife Renée, the fellowship provides tenured, mid-career faculty in the Electrical Engineering and Computer Science Department with resources for up to three years to pursue new research and development paths, and to make potentially important discoveries through early stage research. Since 2003, Prof. Ozdaglar has been a member of the faculty of the EECS Department, as well as a member of Laboratory for Information and Decision Systems (LIDS) and the Operations Research Center. Her research interests include optimization theory, with emphasis on nonlinear programming and convex analysis, game theory, with applications in communication, social, and economic networks, and distributed optimization and control. She is the co-author of the book entitled “Convex Analysis and Optimization” (Athena Scientific, 2003). She also co-directs (with Prof. Sandy Pentland) the Center for Connection Science and Engineering, a virtual center at MIT to bring together network initiatives from across the Institute. See page 17 in this issue of the Connector. Asu Ozdaglar

34

www.eecs.mit.edu

Professor Ozdaglar is the recipient of a Microsoft fellowship, the MIT Graduate Student Council Teaching award, the NSF Career award, Class of 1943 career development chair, the 2008 Donald P. Eckman award of the American Automatic Control Council, and is a 2011 Kavli Fellow of the National Academy of Sciences. She served on the Board of Governors of the Control System Society in 2010. She is currently the area co-editor for a new area for the journal Operations Research, entitled "Games, Information and Networks", an associate editor for IEEE Transactions on Automatic Control, and the chair of the Control System Society Technical Committee “Networks and Communications Systems”. n

Vladimir Bulovic´ was appointed to the Fariborz Maseeh professorship in Emerging Technology in January, 2013. Vladimir is a widely recognized leader in the areas of energy and nanotechnology. The Fariborz Maseeh chair was previously held by President L. Rafael Reif. Vladimir has made pioneering contributions to the fundamental understanding of organic and nanostructured optics and electronics, and has applied his findings to develop devices that define the state of the art. Together with his former students he has founded three start-ups that presently employ over 200 people: in 2005, QD Vision, Inc., which produces quantum dot optoelectronic components; in 2008, Kateeva, Inc., which is focused on development of printed organic electronics; and in 2011, Ubiquitous Energy, Inc., which is developing nanostructured solar technologies. In 2012, Vladimir shared the SEMI Award for North America in recognition of his and his colleagues’ contributions to commercialization of quantum dot technology. Vladimir has also made outstanding contributions to MIT’s energy research and education. He played a critical role in the establishment of the Energy Studies Minor Program, the first program to link all of MIT’s departments and schools. Vladimir’s educational contributions to the EECS department include the development of 6.789 (a graduate class in Organic Optoelectronics); co-development of 6.007 (an undergraduate class on Electromagnetic Energy: from Motors to Lasers); and the newest introductory class, 6.S079 — Nanomaker, which emerged from the freshman seminar he has taught for many years with Prof. Rajeev Ram. Vladimir’s educational contributions have been recognized by the Ruth and Joel Spira Award, the Bose Award for Distinguished Teaching, Eta Kappa Nu Honor Society Award for Outstanding Teaching, Class of 1960 Fellowship, and most recently by the Margaret MacVicar Fellowship.

Srini Devadas was appointed the Edwin Sibley Webster Professor of Electrical Engineering and Computer Science, joining Prof. Alan Willsky as the second Edwin Sibley Webster chaired professor at MIT. For nearly sixty years, many prominent faculty members have held this professorship, including Ernst Guillemin in 1960, Lan Jen Chu in 1963, Peter Elias in 1974, and Ronald Rivest in 1992. Professor Devadas has done pioneering work in a number of areas related to CAD, security and computer architecture. His early award-winning work involved developing a symbolic simulation method for analyzing the average and worst-case power estimation of combinational logic; this was among the first efficient, accurate power estimation methods developed. Professor Devadas was one of the first to recognize that manufacturing variations in integrated circuits could be used to not just identify, but to authenticate, individual integrated circuits. He coined the term Physical Unclonable Functions (PUFs) in 2002; PUFs are now a very active field of research and this technology has been commercialized. Most recently, along with his students, he has developed a computation-migration-based parallel processing architecture where programs move to where the data resides rather than the other way around. As proof of concept, his group is working on the tape-out of a 121-core processor. In addition to his research, his service and teaching record at MIT has been extraordinary. He served as Associate Department Head of the EECS Department for nearly 6 years, leading the Computer Science side of the department during that time. He has taught 6.00, 6.001, 6.002, 6.004, 6.005, 6.006, 6.042, 6.046, and 6.170 at the undergraduate level, and graduatelevel classes in VLSI, architecture and security. n

MIT EECS Connector — Spring 2013

35


Faculty News : Chairs inventions made major contributions to the development of the electrical industry. Dr. Thomson served as a member of the MIT Corporation and its Executive Committee, as a non-resident professor of applied electricity, and in one critical period, 1920-21 as Acting President of the Institute. The Elihu Thomson Chair was previously held by Professors Hermann A. Haus and Erich P. Ippen. n

James Fujimoto was appointed the Elihu Thomson Professor of Electrical Engineering for a five-year term, July, 2012. Prof. Fujimoto joined the MIT faculty in 1985 as Assistant Professor of Electrical Engineering after completing his SB ’79, SM ’81, and PhD’84 at MIT. He is currently a Professor of Electrical Engineering at MIT and an Adjunct Professor of Ophthalmology at Tufts University. As principal investigator in the Research Laboratory of Electronics, working with his group and collaborators, Prof. Fujimoto pioneered the development of optical coherence tomography (OCT) in the early 1990s. OCT is a new medical imaging modality, which uses echoes of light to enable real-time visualization of internal tissue microstructure and pathology. The development of OCT stemmed from the group’s early studies using femtosecond optical pulses to perform optical ranging and measurement in the eye. This ground-breaking research at MIT in collaboration with investigators from MIT Lincoln Labs, the Harvard Medical School and Tufts University School of Medicine, has resulted in a host of valuable OCT applications spanning ophthalmology and cardiology as well as fundamental research. The technology has had a major clinical impact in ophthalmology, with tens of millions of OCT imaging procedures performed yearly and more than 8 companies which develop OCT instruments for the ophthalmic market. Prof. Fujimoto has published over 375 journal articles, is editor or author of 5 books, and co-author of numerous U.S. patents. He is a fellow of the National Academy of Engineering, the National Academy of Science and the American Academy of Arts and Sciences. His many honors include the 1999 Discover Magazine Award for Technological Innovation, the 2001 Rank Prize in Optoelectronics and the Carl Zeiss Research Award in 2011. The Thomson chair was established to honor the distinguished scientist, engineer, and inventor whose discoveries and 36

www.eecs.mit.edu

In January, 2013, Gregory Wornell [SM ’87, PhD ’91], was appointed the Sumitomo Electric Industries Professor of Engineering. Greg Wornell is a recognized leader in the fields of signal processing and information theory. Prof. Wornell’s research lies where information meets the physical world. One important focus has been on the design of reliable high-speed wireless communication networks. He has a long track record of making contributions in this area. As a recent example, Greg and his collaborators have developed a novel ‘rateless’ coding architecture that drastically simplifies and improves upon the current systems embodied in existing RF standards. Greg's interests include optical and acoustic communication as well. For instance, he and his colleagues have developed coding architectures for multi-mode photon-efficient free-space optical communication. And he is co-designer of the recently demonstrated ‘super-Nyquist’ modem for underwater acoustic communication. With a strong multidisciplinary approach to his research, Prof. Wornell has worked on neural decoding from pre-motor cortex measurements in advanced real-time brain-machine interfaces. He has also worked on emerging millimeter-wave imaging system design based on novel computationallyenhanced digital phase-array technology. He has also worked on security techniques that exploit a variety of physical phenomena, a recent example of which is his development of codes for practical entanglement-based quantum secret-key distribution systems.

Faculty News : Career Development Prof. Wornell has made major contributions to the curriculum of EECS, which over the last decade, have redefined and shaped the department's curriculum in statistical inference. Together with Prof. Polina Golland and other faculty, he led the development of two graduate courses in statistical inference, 6.437 and 6.438. These courses introduce graduate students to the fundamentals of inference, and its interactions with the principles of information and computation. The courses are very popular among EECS students and attract a broad set of students from many departments in the School of Engineering and across the Institute. For his contributions, Prof. Wornell was awarded the Smullin Award for Teaching Excellence. Several other universities have adopted his viewpoint on this topic and have utilized the course material as a template for their own classes. More recently, Prof. Wornell and his colleagues spearheaded a new undergraduate course on statistical inference. Currently offered in its pilot form as 6.S080, the course is slated to become an integral part of the EECS undergraduate curriculum. n

Career Development Faculty Appointments

Adam Chlipala

Yury Polyanskiy

Dirk Englund

Daniel Sanchez

Adam Chlipala was selected to hold the Douglas Ross (1954) Career Development Professorship of Software, July, 2012. Prof. Chlipala joined MIT in July 2011 as an Assistant Professor of Electrical Engineering and Computer Science and a member of CSAIL. He received a BS from Carnegie Mellon in 2003 and a PhD from Berkeley in 2007, both in computer science. Prof. Chlipala's research applies computer theorem proving to construct more effective software development tools. Much of his work involves proving theorems about the correct behavior of software using the Coq theorem proving system, about which he has written a popular online book, due out soon in hard copy from MIT Press. He also works in the design and implementation of new programming languages, including Ur/ Web, a domain-specific language for Web applications, which applies theorem proving technology to rule out costly mistakes in orchestrating the many pieces of a realistic Web site. Dirk Englund joined the MIT Electrical Engineering and Computer Science Department faculty in January 2013 as Assistant Professor of Electrical Engineering and Computer Science. He was appointed the Jamieson Career Development Professor in Jan. 2013. Read more about Prof. Englund in this newsletter under New Faculty, page 38. Yury Polyanskiy was selected to hold the Robert J. Shillman (1974) Career Development Professorship of Electrical Engineering and Computer Science July, 2012. Yury Polyanskiy joined MIT on July 1, 2011 as an Assistant Professor of Electrical Engineering and Computer Science, and a member of LIDS. He received the MS degree in applied mathematics and physics from the Moscow Institute of Physics and Technology in 2005 and the PhD degree in electrical engineering from Princeton University in 2010. In 2000-2005 he led development of the embedded software in the Department of Surface Oilfield Equipment, Borets Company LLC. His thesis work initiated a systematic approach to studying impact of finite delay constraint on information theoretic fundamental limits. The accompanying journal paper won the 2011 Best Paper Award from IEEE Information Theory Society. Prof. Polyanskiy was also a recipient of the Best Student Paper Awards at the 2008 and 2010 IEEE International Symposiums on Information Theory (ISIT) and a Silver Medal at the 32 International Physics Olympiad (IPhO) in 1999. Yury's general research interests include information theory, coding theory and the theory of random processes. His current work focuses on non-asymptotic characterization of the performance limits of communication systems, optimal feedback strategies and non-Shannon information measures. Daniel Sanchez was selected to hold the TIBCO Career Development Professorship in the MIT School of Engineering. This professorship is made possible by the TIBCO Founders Fund to advance research and teaching in computer science at MIT. Read more about Prof. Sanchez and his research on page 38.n MIT EECS Connector — Spring 2013

37


Faculty News : New Faculty

Faculty News : 17th President of MIT L. Rafael Reif becomes the 17th President of MIT MIT’s global strategy; promoted a major faculty-led effort to address challenges around race and diversity; helped foster the emergence of an innovation cluster adjacent to MIT in Kendall Square; led the development of MITx, the Institute’s new initiative in online learning; and led MIT’s role in the formation of edX, the recently announced partnership between MIT and Harvard University that builds on MITx and that aims to enrich residential education while bringing online learning to great numbers of people around the world.

Dirk Englund Dirk Englund received his BS in Physics from Caltech in 2002. Following a year at TU Eindhoven as a Fulbright Fellow, he did his graduate studies at Stanford, earning his MS in electrical engineering and PhD in Applied Physics in 2008. He was a postdoctoral fellow at Harvard University until 2010, when he became Assistant Professor of Electrical Engineering and of Applied Physics at Columbia University. He moved to MIT in 2013 as Assistant Professor of Electrical Engineering and Computer Science and a member of RLE and MTL. His research focuses on quantum technologies based on semiconductor and optical systems. Recent recognitions include the 2012 DARPA Young Faculty Award, the 2012 IBM Faculty Award, the 2011 Presidential Early Career Award for Scientists and Engineers, the 2011 Sloan Research Fellowship in Physics, the 2008 Intelligence Community (IC) Postdoctoral Fellowship, and the 2012 IEEE-HKN Outstanding Young Professional Award. n

Daniel Sanchez Daniel Sanchez joined the EECS Department in September 2012 as an Assistant Professor and principal investigator in the Computer Science and Artificial Intelligence Lab (CSAIL). Prof. Sanchez was appointed as the TIBCO Career Development Professor of computer science at the MIT School of Engineering. He earned a PhD in electrical engineering from Stanford University in 2012, an MS in electrical engineering from Stanford University in 2009, and received a BS in telecommunications engineering from the Technical University of Madrid, Spain, in 2007. Daniel is broadly interested in computer architecture and computer systems. His research strives to improve the performance, efficiency and scalability of future parallel and heterogeneous systems, and to enable programmers to leverage their full capabilities easily. His current projects focus on designing parallel architectures that provide quality-of-service guarantees; building scalable and efficient memory hierarchies for thousand-core chips; introducing, exposing, and transparently managing heterogeneity in the memory hierarchy to improve efficiency; and designing dynamic fine-grained runtimes and schedulers using both software and hardware to improve the utilization and ease of use of these highly parallel systems. n

L. Rafael Reif “We are all in this great enterprise together!” newly inaugurated MIT President L. Rafael Reif declared to an enthusiastic crowd under a huge pavilion in MIT’s Killian Court on September 21, 2012. The buildup to the Inauguration proved just as exhilarating as the event itself. On May 16, the MIT News Office posted the announcement of the MIT Corporation’s selection of the Institute’s 17th president. This announcement read in part: L. Rafael Reif, a distinguished electrical engineer whose seven-year tenure as MIT’s provost has helped MIT maintain its appetite for bold action as well as its firm financial footing, has been selected as the 17th president of the Institute. Reif, 61, was elected to the post this morning by a vote of the MIT Corporation. He will assume the MIT presidency on July 2. As provost since 2005, the president-elect has inspired innovation and played a critical role in the financial stewardship of the Institute. As the Institute’s chief academic officer since 2005, Reif led the design and implementation of the strategy that allowed MIT to weather the global financial crisis; drove the growth of

38

www.eecs.mit.edu

Prof. Reif has been a member of the MIT faculty since 1980, formerly the Fariborz Maseeh Professor of Emerging Technology in the Department of Electrical Engineering and Computer Science. Prof. Reif served as associate head of the Electrical Engineering and Computer Science Department from 19992004 and as department head until his appointment as MIT Provost in August 2005. Upon the announcement, President-elect Reif said: “I am deeply honored to be elected president of the Institute I love so dearly," Reif said. "MIT’s impact on my life — how I think, how I make sense of the world, and how I align my personal aspirations with the call to service — has been profound.” Beginning Sept. 19, 2012, pre-inauguration festivities included presentations, symposia, concerts and a day of post inauguration campus-wide activities from the Beaver Dash (5K+ a few smoots) race, benefitting Habitat for Humanity to an “Inaugural Global Barbecue.” n MIT EECS Connector — Spring 2013

39


New Classes in EECS 6.S02, a medical technology alternative to 6.02 (Introduction to EECS II) is launched

diagnosis. Joel Voldman, who is developing a glucose detection module for 6.S02, is teaching students about noise and model-based inference. Undergraduate Instructor Gim Hom developed the ECG hardware used in the cardiovascular module. The freshman/sophomore level 6.S02, along with new junior/ senior level courses in synthetic biology (jointly with BE) and medical device design (jointly with ME), are part of a broader EECS strategy to expand the role of medical technology in the undergraduate curriculum. Prof. Stultz makes note of several reasons for this expansion. First, more than a third of the

EECS faculty are involved in research that is aligned with medical technology and/or clinically relevant questions; second, there are significant numbers of undergraduates who are either keenly interested in medical applications of EECS or who are premedical students in Course 6; and third, we believe that our students, armed with a rigorous EECS background and a significant appreciation of medical applications, will have a major impact on the quality and availability of medical technology, a benefit to both patients and the broader medical community. As 6.S02 gets underway, students are absorbed in the ECG lab. Course 6-2 Sophomore Eann Tuan has just set up the apparatus with her lab partner Course 18 and 6-3 junior Isabella Tromba. Tuan has always been interested in biomedical research and technological advancements in medicine and health care. “The very first day,” she notes, “Professor Stultz mentioned that he was a Cardiologist while explaining the goals of this experimental class. Right away, I was drawn in. After all these years of watching Grey's Anatomy, I was going to learn from a practicing doctor!” Isabella Tromba, a competitive swimmer at MIT, says about her experience so far: “It was pretty neat to observe that my heartbeat is slower than normal, and there is a very small difference from when I hold my breath and when I breathe normally. In the next coming weeks, we will be working with MRI's, something that I am excited to experiment with and see how we can use computational tools to provide new insights into medical diagnoses and treatments.”

Instructor Gim Hom with students in the lab for 6.S02, spring term 2013. [Photos: Patricia Sampson/EECS]

As noted in the MIT Spring 2013 catalogue: This spring 6.S02 (or 6.02M) is being offered as an experimental version of 6.02 to provide a medical technology context for learning fundamental concepts in information extraction and representation. The class explores biomedical signals generated from electrocardiograms (ECG), glucose detectors, and magnetic resonance images. Topics for the new class include physical characterization and modeling of systems in the time and frequency domains; analog and digital signals and noise; basic machine learning including decision trees, clustering, and classification; and introductory machine vision. Lead faculty member Collin Stultz explains that the class is based on 6.02, the second of the two landmark introductory curriculum classes developed in 2008. “A lot of the material that is covered in 6.02 will have a slightly different flavor,” he says, “— a medical-based technology introduction to EECS.” Prof. Stultz provides the thinking behind 6.S02. The idea is that the new class will engage the fundamental principles at play in EECS. From the standpoint of clinically relevant questions,

40

www.eecs.mit.edu

In preparation for the lab classes starting April 25, ten portable MRI’s have been specially designed at the MGH Martinos Center for Biomedical Imaging by a team under the direction of EECS Visiting Associate Professor Lawrence Wald. Team members include MIT EECS graduate student Clarissa Zimmerman, Harvard postdoctoral associates Jason Stockmann and Cris LaPierre, and MGH Martinos Center site physicist Thomas Witzel.

the course begins where students learn about electrocardiograms in detail and then they learn fundamental things within signal processing like Fourier transforms, power spectra, and the application space is the electrocardiogram. They'll be able to record their own electrocardiograms, do some simple analyses of that data, sample electrocardiograms of people who are sick, be able to do some analyses of these, and then use that information to do clinical decision-making, which then falls into the realm of computer science. So they'll expand the spectrum of these fundamental engineering concepts applied to real biological signals and then apply clinical to analyze this material. He notes that this content brings together faculty who are interested in these problems both on the EE and CS sides. The combination of medical technologies and application of machine learning principals in 6.S02 has been possible through the input and instruction by Prof. Joel Voldman, Instructional Staff member Gim Hom and Prof. John Guttag. Prof. Guttag, whose current research is centered on the application of advanced computational techniques to medicine, is teaching machine learning techniques as applied to medical

Top: Clarissa Zimmerman, EECS graduate student works on the portable MRI, which she helped build under Visiting Associate Professor Lawrence Wald at the MGH Martinos Center for Biomedical Imaging. 6.S02 students Eann Tuann, photo left, and Isabella Tromba, photo right, set up and take ECG readings in lab. [Photos: Patricia Sampson/EECS]

With the goal of enabling the 6.S02 students to get an understanding of all the components and the functioning of the MRI system — rather than viewing it as a ‘black box’— the scanners have been assembled as hands-on, low-cost table-top units. The MRI systems have all the essential components of a clinical scanner (magnet, console, gradient coils, gradient amplifier, RF coil, transmit power amplifier and receiver pre-amplifier). With help from colleagues at other universities, the team has been able to build the system using very inexpensive components. The sample size for the scanner is about 1cm3. However, the scanner has sufficient resolution to generate multi-slice images of interesting biological samples, like mouse brains and hearts. n

MIT EECS Connector — Spring 2013

41


New Classes in EECS basic skills in wet-lab synthetic biology, and potentially get involved in research.”

6.S193: Biological Circuit Engineering Laboratory

Prof. Tim Lu, the lead faculty for the class, notes: “Launching a new lab class that intersects experimental and computational aspects of synthetic biology has been an intense challenge involving over six months of preparation. Thus, I would like to thank the dedicated teaching staff from my lab, including Sebastien Lemire, Allen Chen, Jerry Wang, and Michelle Lu, and my fellow instructors, Ron Weiss and Rahul Sarpeshkar, for their hard work with me on this new educational endeavor. In addition, we are excited that our students are enthusiastic to learn about this growing field and to help us enhance the course for future iterations.”n

The teaching team for 6.S193, Biological Circuit Engineering Laboratory, from left includes (seated) EECS Professors Rahul Sarpeshkar, Ron Weiss and Tim Lu and standing behind them, TAs Jerry Wang, Michelle Lu and Allen Chen. [Photo: Patricia Sampson/EECS]

Biological Circuit Engineering Laboratory (6.S193) is an interdisciplinary laboratory class that aims to train students in the emerging engineering discipline of synthetic biology and to equip them with the hands-on experimental and computational skills to pursue research in this area. Synthetic biology is a growing field, which focuses on engineering biological systems to achieve novel functionalities. Ultimately, we plan for this course to be a recurring cornerstone of a new curriculum in synthetic biology at MIT. This year, Professor Tim Lu, Rahul Sarpeshkar and Ron Weiss are piloting the course with the generous support of the d'Arbeloff Fund, the EECS Department, Biological Engineering, the MIT Biology Department, and Quintara Biosciences. There are currently 15 pioneering undergraduate and graduate students from diverse disciplines such as EECS, Biological Engineering, and Mechanical Engineering. Biological Engineering (Course 20) sophomore Edgar Aranda-Michel, notes on signing up for this new offering: “I remember when first getting the email about the class I was

42

www.eecs.mit.edu

puzzled, a course 6 lab that was wet...this has to be some sort of mistake, but sure enough there it was. What really stood out was the course 20 and 6 flavor the class offered. The course 20 side coming from synthetic biology and building new circuits inside of cells and the 6 side coming from the modeling and predicting of these circuits. I enjoy the fusion of working with awesome professors in a small group setting. Right now I work in synthetic biology, however my focus is prokaryotic systems. This class was sure to give me experience in the eukaryotic systems. However, in taking the class I found that it really had much more to offer. I look forward to lab every time we have it.” Exposure to synthetic biology as a new interdisciplinary field is proving one of the big draws to 6.S193. Course 6 senior Rishi Patel notes: “I am taking 6.S193 because to me synthetic biology seems to be an exciting and new interdisciplinary field. As someone with an electrical engineering and physics background, I think it is important to have primer courses like these that introduce non-experts into the area (particularly those with little or no biological lab training). This class provides excellent hands-on opportunities for learning

Left column: Top: Course 6 senior Aakanksha Sarda prepares some bacterial plates in the ‘wet lab’. Lower left: Steven Keating, Mechanical Engineering doctoral candidate in digital fabrication has recently declared a minor in Synthetic Biology because he believes the future of physical fabrication lies within designed biological systems. Lower right: Edgar Aranda-Michel, Biological Engineering sophomore, checks the recently inoculated plates in the wet lab. Right top: From left, Edgar Aranda-Michel, Rishi Patel and Aakanksha Sarda work through a problem set in a 6.S193 lab class. Patel, a joint Course 6-1 and Course 8 junior, appreciates being able to take this lab to learn basic skills in wet-lab synthetic biology for possible research involvement in the future. Middle right column: Lab class students work out problems in synthetic biology. Lower right: Zachary Banks and Kristen Cotner work in the 6.S193 wet lab. [Photos: Patricia Sampson/EECS]

MIT EECS Connector — Spring 2013

43


New Classes in EECS 6.S196 brings out the human side of technology Principles and Practice of Assistive Technology

Titled by the first name of the client, the three-member ‘Project Janet’ team worked at The Boston Home, a not-forprofit, nursing-care for adults with multiple sclerosis (MS) and other progressive neurological diseases resulting in mobility impairment, and one of several Boston-area health-care facilities that partner with PPAT. Project Janet team members Joy Ekuta, a senior in Course 9 (Brain and Cognitive Sciences), Mechanical Engineering graduate student Casey Jianzhi Chiou, and Course 6 student Jason Strauss worked with TBH client Janet Gardner to develop a more accessible nurse call system. The team divided up responsibilities based on a matching of their respective interests orchestrated by the PPAT staff. Coming from a medically oriented background, Joy Ekuta researched and collected data on MS, documented all aspects of the project (now available at the open source site Make: Projects (http://bit.ly/12jmJLr) and was the team’s primary liaison to Janet and her caregivers at TBH. Casey Chiou focused on hardware design and assembly — giving the project a physical, tangible form. Finally, Jason Strauss was responsible for the software that allowed the system to recognize Janet’s voice when she spoke the activation phrase “nurse call,” as well as for making the microphone, buttons, LEDs and nurse-call system interoperate via an Arduino.

Jason Strauss, Course 6-3 senior, holds in his left hand the Arduino (on photo right), an open source microprocessor which he and his teammates programmed to have the microphone, buttons, and LEDs interact with each other and to allow them to trigger The Boston Home's nurse call system. In his right hand (photo left) is an Arduino EasyVR Shield, a board which can be added to the Arduino, which provides the Arduino with voice recognition capabilities. By using off-the-shelf hardware, documenting their design on a makezine blog, and putting their code on github, the team hopes that other people will be able to build and customize their system.

For at least a decade, Professor Seth Teller and members of his Robotics, Vision, and Sensor Networks Group (RVSN) have pursued machine perception for autonomous robotics and human assistance. From self-driving cars and wheelchairs to body-worn navigation aides, Prof. Teller’s lab has developed machines that can share people’s goals — and do their bidding — in a variety of environments. It was thus a natural outgrowth that working with Professor Rob Miller, whose interests are at the interface of programming and human computer interaction, Prof. Teller developed 6.S196: Principles and Practice of Assistive Technology (aka PPAT), first offered in fall 2011. The interdisciplinary class is designed for small teams of students to work with clients living with disabilities in the Cambridge area to develop assistive technology, a device, mobile application or other solution — to enable each client to live more independently. Initial funding was provided by MIT’s Alumni Class Funds.

44

www.eecs.mit.edu

Seth Teller notes the particular appeal of PPAT to students: "This class is a good fit for students interested in public service, user-centered product design, working closely with a client with a disability (potentially in consultation with the client's caregivers and/or clinicians), and tackling difficult, real-world problems." At the close of the fall term, 12 PPAT students, organized into five teams, presented their work to fellow class members and several assistive technology professionals invited from outside MIT. The projects demonstrated a variety of new assistive technologies including augmented caregiver access and 911 capability for a client with ALS; accessible tablet-based control of an adjustable bed for a client with MS; accessible touchand speech-based nurse calls for a client with MS; using machine vision to make an inaccessible coffee-maker touchscreen accessible for a blind client; and a vibrating bracelet to notify a blind and hearing-impaired client of incoming calls on her mobile phone.

problems was exciting and broadened my way of thinking about my own project.“ Joy Ekuta appreciated taking PPAT because it allowed her the chance to actually help someone in the community. She noted: “Being in Course 9, we don't have a lot of hands-on opportunity to work with people who have any of the brain injuries that we study. This class gave me an opportunity to actually work with someone who had a degenerative neurological condition, while also having a great impact. The biggest opportunity I saw with this class was to do something that went beyond the classroom to change someone's life.” Profs. Teller and Miller plan to offer the class again in fall 2013. n

Jason and Casey designed the hardware and software modularly so that more buttons, microphones and even other input devices could be easily added. Using off-the-shelf hardware and software including a voice-recognition application, the team was able to set up an input system to trigger a variety of feedback LED’s around the room, letting Janet know that the nurse call was activated. Jason notes: “Our code is available on github with instructions on how to modify it for similar use cases.” Through multiple lectures and demonstrations offered by PPAT, team members in the class were able to build systems that could be readily adapted for other clients with related, but individual and varying needs. Casey Chiou notes of the class: “Our client, living with MS, is a clear example of someone with mobility impairment, so seeing such alternative devices helped give us new ideas and things to consider when designing our own solution. The sessions also taught us about the user-centered design process and how to properly collect data and test in iterations.” PPAT class members also benefited from frequent interaction between teams. Jason Strauss found it helpful to hear about other teams’ projects and solutions during the term, noting: “Engineering problems at MIT often have a correct answer. The issues that arose with these projects often had peculiar and particularly creative solutions. Seeing the way other MIT students from a variety of majors tackled these unique

Top: Joy Ekuta tests nurse call system equipment with Janet Gardner in Dorchester, Mass. Ekuta is helping to design a new nurse call system for Gardner, who has multiple sclerosis. Gardner's old system used a single button attached loosely to a wall, which she often dropped. [Photo: M. Scott Brauer, courtesy MIT News Office] Lower: Casey Chiou, left, and Jason Strauss, right, work with Janet Gardner to make final adjustments for the nurse call system. [Photo: Patricia Sampson/EECS]

MIT EECS Connector — Spring 2013

45


Staff Features Claire Benoit

Administrative Assistant II EECS Graduate Office

Things were different in the 60’s when I was looking for my first job. There really were not a lot of career choices for women like there are today. Most of us were teachers, nurses, or secretaries. I received my Associate’s degree in Business in June 1965, and spent that summer on the Cape with a couple of college roommates. In the fall, it was time to look for a “real” job. I decided to apply to MIT since I had heard that it was a great place to work. There was a job opening in the EECS Graduate office (EE at that time) working with the Student Administrator, Dot Young. I interviewed with her and Prof. Truman Gray and was hired. Back then to get a job as a secretary you must have passed a typing test at Personnel, before being sent out for interviews. Part of my interview included translating a letter from shorthand that was dictated by Prof. Gray. I remember that he threw in some tricky words. Fortunately spelling was my strong suit (no spell-check to rely on back then) and I’m sure that was part of why I got the job. I worked with Dot Young for a couple of years assisting with student records, and eventually moved over to the admissions side of the office working with first Fred Fairchild and later Horace Smith. I left MIT in 1972 to have my first daughter (subsequently had another daughter) and did not return to the work force until 11 years later. When I returned to MIT in 1983 I worked part time in the Civil Engineering Department which is now Civil & Environmental Engineering. I was an Administrative Assistant to faculty members, the principal one being Jerome Connor. After a few years I became full-time and I stayed in this department for a total of 10 years and did the normal administrative tasks like monitoring accounts, helping with faculty searches, photocopying course work, etc. In 1993 the department went through a major restructuring and I knew my job might be in jeopardy. At

46

www.eecs.mit.edu

the end of 1993 I saw a job listing in Tech Talk in the EECS Graduate Office, again on the admissions side. I jumped at the opportunity to go back to my first MIT home, and found myself working with Peggy Carney who was secretary of the admissions committee.

Agnes Chow

EECS Admnistrative Officer

EECS is the largest department at MIT and the application count attests to this. When I left in 1972, the department received 900 applications for graduate admission. Today the count is close to 3,000. The best thing to happen along the way was the development of the online application by two of our CS faculty members. It has been in effect for close to 9 years, and I don’t think anyone involved in admissions would want to go back to paper applications. It is not a paperless process though. We still require original transcripts even though the applicants upload a copy to the online application. I’m guessing that the transcript count is close to 8,000 per year since most students have multiple transcripts sent in. These transcripts arrive by two venues, postal and courier service. They all need to be filed of course, and this is how we spend Christmas vacation — the biggest filing job ever and we take over the Jackson Room for a couple of weeks. Working at MIT has been a lot of fun — it truly is a great place to work — and at times it can be quite challenging. We deal with many prospective international applicants and it can be difficult at times to respond to the many telephone inquiries without the benefit of face-to-face conversation. Thank heavens for email. When dialogue gets difficult due to poor phone connections, I suggest they send me an email. The changes from my first stint in EECS and my current stint are astounding. Back then we had typewriters, not computers. Our copiers were mimeograph and duplicating machines and depending on which one you used, you were covered with purple or black ink by the end of the day. I used to type theses on the side for extra money. How I managed that I’ll never know. Try to imagine typing a technical thesis on a typewriter with an original and two carbon copies — the Institute requirement at the time. Add to that the tediousness of constantly changing the ‘elements‘ to move from English to Symbol to do equations and Greek letters. It is a totally different story today. Most students do their own typing with no need to hire a typist. What is really inspiring to me are the people I have met at MIT over all the years. I have had the privilege of working with wonderful colleagues and this includes faculty, staff, administrators, support staff and of course students. I will sorely miss the interaction with them when the time comes that I do leave. MIT is a very down-to-earth place (not stuffy like some universities) and that has been the major draw for me. I truly feel blessed to have had the good fortune of working in a worldclass institution like MIT with people I admire and respect. n

Photo by Barbara Chen, 2013

discovered that singing brought laughter to the temporarily parentless household; music drew her and her brothers closer together. Against all odds, Agnes managed to pass the high school entrance exam, and was admitted to the Hsinchu Girls High School, the best high school in town. She knew then that no one could stop her from pursuing her education. Agnes’s father passed away during the summer of her freshman year at Soochow University in Taipei. After receiving her Bachelor’s Degree in Accounting from there and working for a couple of years at an accounting job for the Taiwan Fertilizer Corporation, she came to the United States to study at SUNY Binghamton, where she received her Master’s Degree in Finance and Business Enterprise. Before entering SUNY, however, she worked at the General Latex and Rubber Company, located on Main Street in Cambridge, as an accountant. During that period, she joined the MIT-Harvard Chinese Collegiate Choral Society, whose weekly rehearsal took place at the MIT student center. There she met Joseph Chow, a student pursuing a PhD in optimal control at Harvard, and three years later they were married. Joe spent his career at Lincoln Lab; their children Jennifer and Jeffrey went on to earn their undergraduate degrees at MIT.

Agnes Chow has spent her professional life in finance and administration. As the firstborn child and only sister to her four younger brothers, Agnes has known the value of responsibility, dedication, determination, and independence from a very young age. Those who know her at MIT recognize these same strengths in her today. Born in Shanghai, Agnes lived her first five years between Shanghai and Soochow, sixty miles away, where she enjoyed playing with cousins in her grandparents’ rock garden. Her life changed drastically when her father, an electrical engineer, decided to join the military to fight in Chiang Kai-shek’s army, and ultimately moved his immediate family to Hsinchu, Taiwan. In Taiwan, Agnes took on responsibilities to help her parents. Her first accounting experience was at age 9, when her father asked her to be in charge of collecting the water and electricity bills for the 10 families living in their common building. She learned to organize and track in a little notebook — her very first worksheet! — the money paid and owed per family, by room number. Life was not easy for Agnes’s parents in Taiwan during those early years. When her mother was ill and required lung surgery at a hospital in Taipei, Agnes was asked by her father to stop school, and not to register for the high school entrance examination, because her help was needed at home. During the month that her parents were away, Agnes took care of her brothers, ages 1 to 10. She gained life-long lessons during that period, about the importance of planning, organization, time management, and the skills to survive in a seemingly impossible situation. She also

Photo by SDK Photo & Design, fall 2012

Agnes enjoys creative thinking, particularly in designing and building administrative structures and processes that facilitate the work of others in the enterprise. She has had the opportunity to work as a profitability analyst at First National Bank of Boston, as a cost analyst at State Street Bank, as a pricing analyst at the Chevrolet Division of General Motors, and as “Island Auditor in Charge” for Global Associates at Kwajalein, Marshall Islands. In January 1984, Agnes started her career at MIT in the Laboratory for Computer Science (LCS). From Staff Accountant to Fiscal Officer, she learned the ins and outs of research administration at

MIT EECS Connector — Spring 2013

47


Staff Features LCS, and benefited from attending budget meetings with the towering and visionary LCS Director, Professor Michael Dertouzos. In 1987 Agnes moved to the Center for Technology Policy and Industrial Development (CTPID), where she worked as Administrative Officer (AO) for Professor Daniel Roos. During her eleven years at the Center, she gained experience in global industrial consortium management and interdisciplinary research management, and also witnessed the Engineering Systems Division (ESD) grow from the white-paper stage to its final formation.

Janet Fischer

Graduate Administrator EECS Graduate Office

In 1998, Agnes moved back to the computer science part of MIT, working for Professor Rodney Brooks at the Artificial Intelligence Laboratory (AIL). Four years later, when AIL and LCS merged into CSAIL, Agnes was responsible for the financial part of the merger. This gave her a unique opportunity to exercise her creative talents — all the way from the financial system infrastructure to the final financial processes. She regards this experience as one of the highlights of her MIT career. At CSAIL, she worked for both Professors Rodney Brooks and Victor Zue. For her work at CSAIL, Agnes was presented with MIT’s “Going Above and Beyond” Excellence Award in 2004–2005. When the EECS Department needed an AO at the end of 2004, Agnes was alerted to this opportunity. She signed on with the department head, Professor Rafael Reif, and moved her career from research administration to academic administration. This is her ninth year at EECS. She has served three department heads (Professors Rafael Reif, Eric Grimson, and currently Anantha Chandrakasan.) She says about this experience: “I have come to appreciate each leader and his leadership team. I have been inspired by their vision, passion, and dedication to building a strong EECS Department in education and research, and by their attention to nurturing the faculty.” In 2010 Agnes was awarded the School of Engineering Ellen J. Mandigo Award for Outstanding Service. She credits the capable EECS staff for their key role in keeping the administrative engine of MIT’s largest department running effectively and smoothly. She is also appreciative of the support and guidance she has received over the years — and continues to get – from the SoE Dean’s office, noting that her career at MIT owes much to SoE Assistant Deans Donna Savicki and Sheila Kanode. During her 30 years at MIT, Agnes and her husband have had the opportunity to travel widely — the Silk Road of China, Abu Simbel in Egypt, Machu Picchu in Peru, as well as amazing European cities and coastlines. But in her heart the highlight was her return visit a few years ago to Shanghai and Soochow, where she started her life’s journey. She and Joe plan to take their children, their children’s spouses Tom and Suzanne, and their four adorable grandchildren — DJ, Sienna, Zac and Zoe — back to the garden in Soochow, where Agnes left her footprints many years ago. n

48

www.eecs.mit.edu

certainly more challenging, but I strive to have positive and meaningful encounters with all the students that come by the office, or contact me in other ways.” In the last couple of years, Janet has enjoyed leading seminars for women and first year students with EECS Graduate Officer, Prof. Leslie Kolodziejski. “This has helped to break down our large department a bit – allowing us to get to know small groups of students very well,” she adds. “The students I meet on an individual basis are brilliant and the best thing I can do is help them with “administrivia” and then get out of the way and let them do their own thing!” Janet says it took her a while to adjust to the EECS culture, after having been in ChemE, where there seemed to be more bureaucracy (in a good way). “I have come to appreciate that there are not a lot of hurdles that EECS grad students have to jump through (although they may not realize that)!” she says. “I have really enjoyed working with our Graduate Officers, first Terry Orlando, and now Leslie Kolodziejski, and the entire Grad Office team and Headquarters staff.”

When Janet Fischer was growing up, she was always interested in the ocean, and, in high school she was quite keen on becoming a marine biologist. “That dream dissipated during freshman year Biology 101 class!” she admits. “I am still interested in the ocean, but it is pretty much limited to walking along the beach these days!” Wheaton College, at the time Janet attended, was a small (1,200 students) women's college, and she was really interested in the residential life programs, and helping fellow students in the dormitory system. As a student manager in the dining hall, she took delight in loading the giant (Hobart) dishwasher with great efficiency. In her senior year, she was a class officer, and got a small taste of leadership through that experience. On graduating from Wheaton with an initial interest in banking and finance, Janet worked at a small financial planning company in Harvard Square. Since 1986, she has worked at MIT. Every position that Janet has held at MIT (with the possible exception of her time in the Provost's Office) has been about customer service. She says: “It is pretty simple — I see my role as helping people reach their goals, and for the most part that is about helping students meet their degree objectives, so that they can then go out into the world and do amazing things!” Beyond that, Janet really hopes that ‘her’ graduate students enjoy the experience called graduate school. When she was in Chemical Engineering (1991-2000), she got to know the student population very well. (It was easier to do with 225 students.) “With 650 students,” she notes, “it is

Some things that Janet has found fulfilling while in EECS include getting into social media (blogging and Facebook), overhauling the qualifying exams, and (at last) getting to do registration and grading ON-LINE—for which she credits the great work by the Registrar's Office and IS&T. She particularly likes crafting the daily announcements, which she emails to all the grad students every morning with a selection of timely opportunities and events. She also enjoys all the parties and events put on by the department, and our student groups, such as GSC, GSA, GW6. Janet values her connections with people at MIT saying: “At MIT the most meaningful thing for me has been getting to know such wonderful people through the years, many of whom are lifelong friends. There are too many to mention!” Janet’s father, Jack Fischer, is a member of the Class of 1959. “He is still an active alum, and I got to assist his class at their 50th reunion, which was really a once in a lifetime experience.” Janet and her father had a lot of fun celebrating the 2011 sesquicentennial, as well as President Reif's inauguration in September 2012. Father and daughter also enjoy following the MIT basketball teams — taking in a few games each year. One of the most rewarding aspects of being Graduate Administrator for Janet is seeing students overcome major hurdles in their lives to ultimately achieve their degree. When Janet worked for the Provost, she found it extremely satisfying to be able to develop a few programs that are still in place and going strong today (Presidential Fellows Program, CONVERGE, Web Grad Aid application, Postdoc Association). She was recognized with the MIT Excellence Award in 2007.

Not one to miss opportunities to meet with students worldwide, Janet has also been involved in the Hosts for International Student (HISP) Program since 2002. She has hosted students from Zimbabwe, India, Kenya, Peru, Singapore, Ecuador, China, Taiwan and Poland. If you haven’t met Janet Fischer, you must introduce yourself! n

Myron “Fletch” Freeman EECS Manager, Educational Computing Facility

Myron “Fletch” Freeman: Self portrait.

Myron Freeman has been known as Fletch (since the movie by that name came out in 1985). He grew up in the smallest county on the Eastern shore of Maryland in a town called Chestertown. Since his exposure as a second and third grader to the Apple II computers in his elementary school library, he has been fascinated with computers. “I just sort of had a love for computers,” he remembers. “I was amazed at the computer and knew that I wanted to do something with computers.” And he did. Fletch took programming courses along the way and into high school, where with friends he would help configure Macintosh computers donated by local companies. He and his friends also helped people get started using them. Fletch wasn’t expecting to get into MIT, but he did and came to MIT as a student for two years. Although he admits he didn’t have his heart set on coming in the first place, he became a dual math and computer science major known at the time as 18C. He also made some life-long friends in the EECS Department Computing Facility. As he was deciding to leave MIT (in 1992), his friends convinced him to stay on. By 1996, after

MIT EECS Connector — Spring 2013

49


Staff Features working in the systems area of the Microsystems Technology Laboratories (MTL) for several years, Fletch started working full time with the EECS Department ultimately becoming Manager of the Educational Computing Facility (ECF) in 1999. “We handle a bit of everything,” he says about the work done by the ECF. This “bit of everything” includes the computing needs for all in the headquarters and the teaching staff on the fifth floor. If any help is needed getting set up for Athena, Fletch and his team handle it and they handle all the lab renovations. “I do all sorts of things, so I get asked all sorts of questions,” he notes. “So, for example, I need to know where the circuit breakers are for the Grier Rooms.” His responsibilities over the years have expanded. Now he takes care of the computing needs in the lab down in the basement of the Stata Center. He notes, “It is isolated and hard to keep track of, so it helps to work with the systems folks in the affiliated labs.” Fletch works closely with the TIG (The Infrastructure Group) in the Computer Science and Artificial Intelligence Lab (CSAIL), the LIDS systems folks, Dave Foss in the Research Laboratory of Electronics (RLE), the MTL systems group and of course the systems people in the MIT Information Services and Technology (IS&T). Since 1996, there have been many system updates across the Institute. Fletch has been involved with all of the updates for the EECS Department. He notes about this history: “We’ve actually done ‘forklift’ updates of the networks where we replaced all the hardware, rewired all of the floors to handle newer networking standards. We’ve obviously upgraded all of the various systems for not only the network but the administrative systems, the data bases, the web servers and the full gamut of everything else at all connected with these needs.” Fletch and the ECF crew coordinated the installation of the

Student Groups new Voice-Over IP (VOIP) phones and subsequent training of all the staff in 2012. One of the more unusual roles that Fletch remembers was taking part in the process of setting a new identity and web presence for the EECS Department in the fall of 2011 through spring of 2012. He enjoyed being part of the creative process, noting: “I don’t normally get to be in on the creative side of things in EECS. I usually just keep the nuts and bolts down. I don’t usually get to sit down and have a voice in some of the creative directions in how the website should look and feel and how we should interact with it. It was lots of fun.” Fletch is also an avid, self-taught photographer. While growing up, he liked photography, but didn’t get involved until he got a used, film Single Lens Reflex (SLR) camera in 1994. Although he took some photos, his interest in photography was fully sparked when photography went digital. “With digital cameras I was not counting the frames,” he says. “I could really learn that skill set and find my own voice in photography.” Other than a few classes at the New England School of Photography, where he enjoyed being a TA for a few night photography classes, Fletch continues to shoot photos that he has submitted for display at MIT and elsewhere. His photos have appeared at MIT for the Campus Activities and Artists Behind the Desk events and displays along the Infinite Corridor and for the MIT Excellence Awards showcase. In 2003, Fletch was recognized for his service with the MIT Excellence Award for going above and beyond providing exceptional client service. If you get a chance, remember to ask Fletch for his favorite recipes. He might tell you how to cook blue crabs – the Maryland way. n

The MIT Formula SAE Team Goes Electric! MIT Formula SAE (Society of Automotive Engineers, http:// www.sae.org/) went through a transformation in 2012-13— gaining a new kind of ‘charge’—as the team decided to switch from an internal combustion engine car to an electric formula racing vehicle. Starting in June 2013, the team will be competing in the first Formula SAE Electric event in a newly designed and built electric racing vehicle. The completely new set of rules and the change in design and need for entirely new powertrain technology has brought a new level of demand for the skills brought by EECS majors. Thus, the MIT Formula SAE flyer with the heading “The Electrical Engineering Perspective” was earnestly distributed to EECS students this past fall. The result is an unprecedented nine EECS students now participating in a group that previously attracted mostly mechanical engineering students. Course 6 senior, Brian Sennett, the FSAE electronics team lead, joined the group in the fall of 2011 when several of his MIT friends on the team recruited him as an EE. He explains that the team calls itself MIT Motorsports – though it is the Formula SAE (FSAE) team from MIT. He says the decision to go electric was in part influenced by the automotive industry’s move away from the internal-combustion drive over the next several decades – influencing the start of the electric vehicle competitions worldwide over the past few years. Sennett is pleased to note that MIT Formula SAE offers lots of take-aways to Course 6 students. “To start,” he says, “we host UROPs during IAP for credit and during the summer for credit or pay. I've also seen the project classes I've taken (6.115/6.131) come through very, very strongly in the technical material I'm working on with the team, and I’ve been able to build working components with those skills. While those project classes have a lot of hands-on, being able to build that into a true product that must function in the real world is invaluable.” Participation on the MIT Formula SAE team also offers built-in leadership skills from having to lead people in putting out a product on schedule, to spec, and to budget. Sennett notes: “Our team echoes a lot of the same processes that we'd see as interns or full-time employees in corporations, and in fact, putting Formula SAE Team on my resume has drawn praise (and often job offers) from most companies I've interviewed with.”

Hooding 2012: The annual EECS Hooding Reception is a time for students, alums, faculty and staff to reconnect and celebrate.

50

www.eecs.mit.edu

With two main roles, Brian Sennett has spent a lot of time designing the structure of the electronic systems in the car (control, interface, safety, data acquisition) and managing the tasks required to get those systems built. He has split up the

Top: The main setup is the dynamometer with the two motors and the torque transducer (pictured) in the middle. This setup will allow the team to calibrate the torque, and energy control. Middle: Course 6 students who are FSAE members (left to right: Brian Sennett, Rachel Luo, Erik Johnson) observe the performance of vehicle control components in a test setup. Bottom: The electric vehicle's main computer, left, and two CANbus "nodes" that relay data to the computer.

MIT EECS Connector — Spring 2013

51


Student Groups Formula SAE Team continued sub-projects such as the vehicle computer, testing the safety system and designing a dashboard LCD with other EECS FSAE members such as Erik Johnson, Course 6 sophomore, and one of the drivers. Course 6 team members have in particular been involved with the "brain" unit, for which many possible microcontrollers are available. Seeing that other teams have chosen from a simple embedded micro to a full Linux computer, the team decided to use National Instruments' sbRIO-9606 single-board computer running LabView. Besides the fact that LabView is designed to flawlessly run machinery and integrate complex sensor systems, it is an excellent programming tool for the block diagrams needed to visualize the car's behavior. Course 6 Rachel Luo has taken on the setup of the LabView system as well as coordinating the team’s website – although she and the other Course 6 team members agree that even now someone with a strong CS background could play a major role in designing and debugging the vehicle control software. Erik Johnson the ‘youngest’ of the Course 6 FSAE team members is pleased with the multiple learning opportunities. He says: “Personally I am interested in Motors and motor controllers so I have been able to apply a lot of reading to using expensive high quality equipment and the controls involved.” He also notes that with the switch to electric car, a lot of extra safety rules are required for design and failure mode analysis.

6.270: The Course 6 autonomous robot competition – entirely run by students Since 1987, Course 6 students have run an IAP course called 6.270, that allows any MIT student to design and build a robot, which will ‘play’ in the end of January competition. The goal of the class is to design a machine that will autonomously navigate around a playing surface, recognize other opponents and manipulate game objects. What sets 6.270 apart from Course 2’s 2.007, is the total lack of human intervention once the competition rounds begin. Through lectures and some knowledge of software (Python), the students who take 6.270 will learn about hardware, software and the information that is needed to design, build and debug a working robot. As noted on their website, “there are no formal prerequisites for 6.270, but they help.” Teams are composed of two or three students and each team is given the same kit containing sensors, electronic components, One of the 6.270 robots successfully unloads its payload. Photo by Christopher A. Maynor—THE TECH

batteries, motors, and LEGO. After an intensive first week of lectures, the remaining three weeks are left for the teams to create a working robot.

Also part of the Powertrain team, Johnson has worked on selecting the motor and the motor controller based on models and constraints. The powertrain group is working on the design of a schema for traction control and torque vectoring – a kind of electronic differential – to enable the dual-motor drive for the car to improve performance. In addition, Johnson has built the dynamometer that the team is using to test various operating points and conditions such as torque control, regenerative breaking, transients, energy tracking, and temperature monitoring.

Among robot competitions, 6.270 also stands out because it is entirely run by students. In what amounts to a year round cycle of organization, the 6.270 student organizers meet from early spring term through the summer and following fall terms to determine the overall approach and detailed planning needed to run the lectures, labs and equipment for the class to proceed through January of the following year. The 6.270 student organizers are generally the students who have taken (and loved) the class and really want to contribute to how it is run. Isaac Gutekunst, this year’s president and vice president in 2012, remembers taking 6.270 as a Freshman in 2009. “My team did not do great,” he says. “We did quite poorly because of a lot of interesting things that I have learned to appreciate over the past few years — not necessarily due to technical ability — but due to lack of knowledge of what was important.”

Other Course 6 members of the MIT Formula SAE team include Marcelo Polanco, LabView system programming; Alan Ho, LabView system programming; Conner Fromknecht, driver interface and LCD control; Ariana Eisenstein, driver interface and LCD control; Brian Alvarez, wiring and electronics packaging; David Kim, wiring and electronics packaging. n

Top: The crowd gathers in anticipation of the 6.270 Competition in 26-100. Bottom: Members of the winning Team – ‘Legolas’. Photos by Christopher A. Maynor—THE TECH

52

www.eecs.mit.edu

For the past two years, 6.270 has doubled (to at least 8) the number of lectures in the first week and also included three 2.5 hour lab sessions to bring everyone up to par on the entire process of wiring the robot and how to make a program and get it onto the robot. Gutekunst notes that increasing the lectures has allowed for all levels to go beyond the basics. He and his organizers also decided to add a faculty advisor to help for extra support and added continuity between leadership transitions. They are pleased that Professor Seth Teller continues to fill this role. The 6.270 organizers are also grateful for the sponsorship support received this year. Sponsors included Analog Devices, ARM, Dropbox, Intempc, Lego Education, MIT CopyTechnology Centers, Oracle, QRST’s, Sparkfun Electronics, and the EECS Department. Apple provided the prizes for the winning team members. Team 10, Legolas, came in first — including John F. William ’16, Laura Jarin-Lipschitz ’16, and Jacob F. Tims ’16 pictured on page 52. Coming in second was Team 24, Yo Yak. Team 22, No Prescription, of Theta Xi, came in third place and Team 13 placed fourth. n

The relevant lesson in taking 6.270, Isaac notes, is to understand the limits of your resources — particularly time — and avoid building a robot that is theoretically perfect, but in practice doesn’t work!

MIT EECS Connector — Spring 2013

53


Student Groups

Alumni Features Deborah Estrin, SM ’83, PhD ’85

“MIT instilled in me an expectation for passion and impact; one that I certainly had early exposure to in my parents, both EE PhDs (Univ Wisconsin, Madison), but at MIT it was ubiquitous.”

This fall has marked the second year for the Undergraduate Student Advisory Group in EECS (USAGE). Created in 2011-12 as part of the EECS department's strategic planning process, USAGE provides critical student input to the department leadership group to help guide curriculum development and enhancements. One important initiative that came out of last year's USAGE group was the new SuperUROP, a year-long advanced undergraduate research program. USAGE team members thought about what they wanted to see in their department, polled their peers, and developed a program that addressed their desire for developing enhanced research skills. As a result, 86 undergraduate students are participating in SuperUROP this year, performing graduate-level research over a wide range of research areas. USAGE was instrumental in ensuring that the program was structured to meet the needs of students interested in pursuing entrepreneurial, industrial or academic career paths. EECS Department Head Anantha Chandrakasan notes, "The input we received from USAGE members in 2011-12 was invaluable in creating the SuperUROP program. I'm excited to be working with this year's group as we create new programs and reshape current ones."

54

www.eecs.mit.edu

The 2012–13 USAGE group is comprised of over thirty students who meet regularly with Prof. Chandrakasan, Prof. Dennis Freeman (EECS Undergraduate Officer) and with Undergraduate Administrator Anne Hunter. Additionally, they meet with the Associate Department Heads - Professors Munther Dahleh and Bill Freeman and other members of the department leadership. They are providing input on a range of issues including curriculum (e.g., a medical EECS program), improving response rates on course evaluations, the role of undergraduate students in faculty search, and IAP activities. USAGE2012-2013 members include Ishwarya Ananthabhotla, Joshua Blum, Moyukh Chatterjee, Deborah L. Chen, Jessica Chen, Stephanie Chen, Cody Coleman, Owen Derby, William Gaviria, Gustavo Goretkin, Bianca S. Homberg, Kevin Hsiue, Alexandra Hsu, Sebastian Leon, Andres H. Lopez-Pineda, Noel Morales, Manushaqe Muco, Santhosh Narayan, Catherine Olsson, Anvisha Pai, Victor Pontis, Devon J. Rosner, Aakanksha Sarda, Denzil Sikka, Nitya Subramanian, Christopher Tam, Jelle van den Hooff, Luis Voloch, Cassandra Xia, Xinyue (Linda) Ye, Yao (Rebecca) Zhang, and Xianzhen Zhu. You can read more about each of them at: http://www.eecs.mit.edu/news-events/ announcements/usage-2012-13-undergraduate-student-advisory-group-eecs n

On June 28, 2012, Cornell University announced that Deborah Estrin had accepted the position of professor of computer science — the first hire for Cornell Tech , the new technology center on Roosevelt Island off Manhattan. As the Founding Director of the Center for Embedded Networked Sensing (CENS) 2002–2012, and a professor of computer science at the University of California Los Angeles (UCLA), Prof. Estrin is noted as a pioneer in networked sensing, using mobile and wireless systems to collect and analyze real time data about the physical world.

late father Gerald Estrin, a professor in computer science at UCLA, was noted for developing reconfigurable computing while working in the von Neumann group at the Institute for Advanced Study at Princeton. Their mother, Thelma Estrin, also a professor of computer science at UCLA, has done pioneering work in the field of biomedical engineering. In familial line, Thelma was inducted into the Women in Technology Iternational (WITI) Hall of Fame in 1999, followed by her daughters Judy in 2002 and Deborah in 2008. Margo Estrin is a medical doctor in California.

“Her forte is building real systems that solve societal and industrial problems,” Charles M. Vest, president of the National Academy of Engineering and MIT president emeritus noted for the Cornell Tech press release. Enthusiastic about this new direction, Prof. Estrin says: “The vision articulated by the founders of Cornell Tech is a perfect match for my interests. Their entire campus will focus on technology innovation, application and impact through both commercial and social entrepreneurship. It is an opportunity to build an institution of teaching and research that engages ‘The City’ as co-innovators.”

As a young girl, Deborah Estrin loved taking an experimental six-year math course in middle and high schools. She was so devoted to this class that when her parents took her with them on a sabbatical to Norway, she stuck with these studies. Family photos reveal her buried in her math book strikingly framed by the Norwegian Fjords. She says about herself as a 12 year old: “I think mostly I was just determined. I took myself seriously at a young age. I think that more than anything else is what helped me.”

In 2007, when Deborah Estrin spoke on receiving the Anita Borg Institute’s Women of Vision Award for Innovation, she credited her family, saying, “I grew up surrounded by the ideals of pursuing science and engineering as a stimulating and creative way to have a positive impact on the world.” Deborah and her two older sisters were raised by two electrical engineers, who she notes were also strong feminists. Their

Deborah Estrin also considers herself very fortunate to have met influential women mentors while she was an undergraduate at Berkeley (BS ’80). Dr. Barbara Simons, noted for her work on electronic voting, was a graduate student there when Estrin was a freshman. Dr. Simons and Sheila Humphreys, then Associate Director of the Women’s Center at UC Berkeley, had convened a group of women in computer science and invited Deborah Estrin to join them. Inspired by this group and the spirit of “Berkeley in the ‘70s”, she notes on graduating: “I left MIT EECS Connector — Spring 2013

55


Alumni Features Berkeley not only with a desire to invent things but also with a lot of idealism and activism. I wanted to fix the world, not just solve technical problems.” Following Berkeley, Deborah Estrin enrolled for graduate work in both the Technology Policy Program (TPP) and computer science at MIT. After earning her masters in 1983, she returned to technical design because, she describes, “I found the nature of social science research very far from social activism. I have a more active than 'pondering' personality. Creating technology is 'active'— it's about doing, about movement.” Working towards her PhD in computer science, Deborah Estrin says MIT introduced her to a world of passionate technologists. Through example and active inclusion, she notes, Dave Clark introduced her to the nascent Internet research and engineering community. She credits her advisor, Jerry Saltzer with taking her as his student despite her mixture of interests. “He believed in me and had high expectations and plenty of constructive criticism!” She also notes that both “…Jerry and Dave valued impact, and system-building and use, over publications.” In the late 80's, soon after graduate school as an Assistant Professor of Computer Science at University of Southern California, Deborah Estrin plunged into collaborative activities in Internet protocol design. Dave Clark introduced her to the Internet Engineering Task Force (IETF) through which she became involved with colleagues at the Information Sciences Institute (ISI). ISI had built and launched the key infrastructure for the Internet—the Domain Name System and the Request for Comments (RFC) process—providing the foundation for today’s Internet economy. She became full professor in 1998. In the late 90's as a member of DARPA's Information Science and Technology (ISAT) study group and surrounded by people pursuing bold new ideas, Prof. Estrin took the initiative to start off in a new direction—wireless sensing. In 2002 Deborah Estrin, by then a faculty member at UCLA, founded the NSF Science and Technology Center for Embedded Networked Sensing (CENS 2002-2012), which she directed for ten years. Through the work at CENS, new technologies in support of environmental monitoring applications were developed and set in place. Deborah Estrin notes that her motivation for wireless sensing was inspired while she was on vacation in Costa Rica. She says: “This sounds like something for the press but I was sitting in the Costa Rican rain forest thinking about how to help ecologists understand and protect such a dense ecosystem where there was so much biodiversity to capture/measure within what would have been a single pixel of a satellite image. Environmental monitoring as the killer app for distributed sensing all started there for me.”

56

www.eecs.mit.edu

Over the past few years, Deborah Estrin has increasingly turned to mobile health. She co-founded with Ida Sim, Professor of Medicine at University of California San Francisco (UCSF), the nonprofit Open mHealth in 2011. They co-authored a position paper in 2010 for Science Magazine putting out the call for open mobile health architecture. Subsequently, the two convened a group of experts from the software and health worlds to bridge health and technology. The online entity Open mHealth invites collaborators from developer resources to health entrepreneurs and professionals to share resources. See: http://openmhealth.org/.

Dropbox Founders Drew Houston ’05 and Arash Ferdowsi '08 Growing up with code and dreaming of startups When Drew Houston was very young, his parents bought a PCjr — their first computer. Soon after, his father, a Harvard University trained electrical engineer, introduced his fascinated five-year old son to programming in BASIC. Although he played his share of computer games while growing up, Drew was far more intrigued by how the games worked. In fact, he taught himself C by reading the source code to a text-based online game.

She notes about this work: “First with colleagues at UCLA (William Kaiser and Gregory Pottie) we started doing mobile sensing since fixed point sensing has limitations of scalability and economics. And around the same time mobile phones were becoming prevalent and Nokia invited me to a workshop. These two things happened to be resident in my head at the same time, and ever since then I have been looking at mobile phones as sensors, as sources of data – and from that to participatory sensing [for both civic engagement and STEM education] to mobile health was really just an obvious progression.”

By freshman year in high school, Drew gained access to a computer game by signing up to test it. Not so interested in playing the game, he instead discovered a variety of security-related bugs and made the game company aware. Soon after, Drew at age 14 was invited to work with them — though his father had to sign the legal documents.

Now in her latest career move as Professor of Computer Science at Cornell Tech (and Professor of Public Health at Weill Cornell Medical College), Deborah Estrin is in a position to pursue these interests from a fresh perspective and environment. Under Cornell Tech ’s mission for technical excellence with a focus on collaborative projects, industry mentors and entrepreneurship and business, Prof. Estrin will not only enhance that mission but also gain new ways to pursue her life goals for creating technology that has a positive societal impact. n

“She really is trying to address problems that matter and that will impact humanity now. As a role model, ...seeing such a strong woman achieve so much means a lot in a field where there aren’t that many women.” — Dr. Nithya Rmanathan Researcher, Computer Science, UCLA (former PhD student with Prof. Estrin)

The idea of innovation was at Drew’s core from an early age. At a college prep talk to his eighth grade class, the speaker asked: “Do you know what you want to do when you grow up?” Drew was the only one to raise his hand. He notes, “I knew that I loved computers and that I wanted to start a company.” He was also a big fan of Bill Gates. Like his Dropbox co-founder, Arash Ferdowsi began programming at an early age. His father introduced Arash to Q Basic while he was in elementary school in Kansas City. While he was in middle school, he took classes at a community college to learn C++. He too was passionate about computers and programming and realized that MIT was the best place in the world to become a computer scientist. He says about this decision, “There was no doubt in my mind that I needed to end up there.”

Drew Houston (’05) and Arash Ferdowsi ('08)

The Ultimate Incubator For Startup Training

were undergraduates in Course VI, com-

Unlike high school, at MIT Drew found he could devote himself to subjects that he was interested in. “First and foremost,” he notes about his MIT Course VI education, “I became a better engineer at MIT. Growing up programming was important, but paired with the theory, I became better much more quickly.” He also came to understand that good business decisions in technology companies are often made by good engineers. “You have all these examples of technologists who have picked up the business side on the job but I can’t think of any examples in the other direction,” he notes.

puter science, at MIT, but they didn’t really meet until the summer of 2007 when it came time to develop Dropbox. What made their Dropbox partnership possible?

MIT EECS Connector — Spring 2013

57


Alumni Features college, where everyone is excited about and dedicated to collaborating towards the Dropbox product goal – to make Dropbox users happy.

One of Arash’s favorite classes at MIT was 6.046, Introduction to Algorithms. In his words: “I love algorithms — I am obsessive about performance and doing everything possible to shave milliseconds off code runtime. 6.046 helped formalize a lot of my intuition around performance and gave me a much deeper understanding of how to design algorithms that truly scale. This was particularly important because my main focus in the early days of Dropbox was to figure out how to make our backend infrastructure horizontally scalable.” Drew wasn't expecting it, but his experiences living in his fraternity at MIT became his first training for managing Dropbox. Not only did he learn to live with and appreciate many people who were naturally gifted in diverse areas, but he also took on offices such as rush chair. Recollecting this responsibility, he says: “You’re handed a budget and some unpaid volunteers, and you've got to make something very complicated happen, which is very similar to my experience with Dropbox.” Although Arash had felt for a long time that one had to have a business degree or a lot of experience to start a company, he became aware through friends in classes and at his East Campus dorm that this was not necessarily true. Besides his Dropbox experience, he credits the “infectious curiosity for hacking and innovating” that he found all around him at MIT. With several friends who ultimately joined Dropbox, Arash started a book exchange website called BookX@mit. At MIT, Drew surrounded himself with people who were also interested in starting companies. He joined the Entrepreneurs Club, took courses like Prof. Hadzima’s ‘Nuts and Bolts Business Plans’ during IAP, and met a lot of other MIT entrepreneurs. He also remembers and values talking with EECS faculty — such as Prof. Charles Leiserson — about their experiences starting companies. Drew has always been interested in both business and management, so it wasn’t surprising that he would take a class or two at Sloan. He describes one in particular. “One of the most eye-opening classes I took was the negotiation class at Sloan. Here was a subject that at first seemed totally opaque to me. Were we going to practice bluffing and yelling?” He discovered well-established and logical frameworks that could allow anyone to become an effective negotiator. He notes: “The important lesson for me was that some things that I didn’t understand at all — leadership, public speaking, management — could be learned and weren’t as mystical as I thought.” Arash’s passion has always been in scale – making things work across thousands of computers and distributed systems that make every aspect of a design horizontally scalable. After his sophomore year at MIT, he applied to intern at Facebook. He notes: “The fact that it was growing so rapidly, meant that

58

www.eecs.mit.edu

Arash devotes the rest of his time to making sure that the product they release works really well and is easy and enjoyable to use. The technical problem solving to achieve this requires a rigorous approach. All Dropbox members meet every Friday to collaboratively review the important developments. During the week, the open layout of the workspace – all on one floor – enables maximum transparency. User operations personnel, who keep on top of user-experience, are embedded with engineers for immediate feedback.

Drew Houston ’05 they were doing pretty innovative things to reach scale. That definitely appealed to me.”

The Right Amount of Frustration, Good Luck and Timing Not long after graduating from MIT and on his way to New York City, Drew experienced a reality check that has now been labeled the ‘aha moment’ for creating Dropbox. He had become comfortable with the MIT Athena environment, where backing up his workstation or forgetting a thumb drive was never a worry. After having graduated, however, he had to upload his entire Linux development environment to a thumbdrive, so he could work at multiple computers (on another startup). Leaving the thumbdrive behind as he boarded the bus for NYC tipped the frustration balance. With nothing else to do, he spent the trip writing out the first lines of code, which would ultimately become Dropbox. That was the summer of 2007 and Drew was told by a venture capitalist to get a partner to build Dropbox. He recalls about this process, “That's one of the great things about MIT — you know what real talent looks like.” That talent was Arash Ferdowsi who Drew describes as “… really smart and yet sort of crazy enough to jump in on this.” Naturally a bit reserved and cautious, Arash notes, “These things don’t generally work out so well.” But in less than five hours, they had decided to go for it. Arash says, “I think part of it was just having the itch like Drew did.” He notes that MIT is really good about this process. “If you are doing well [in your classes], you can take the time to try something, and if it doesn’t work out you can always come back.” He also credits MIT with giving him the confidence to make this decision.

Arash Ferdowsi '08

The Wisdom of Yin and Yang Although Arash did not have startup experience, Drew found that Arash’s talents were refreshingly complementary to his own and that they shared similar intuitions about technology. In running Dropbox, Drew and Arash have similar instincts for how to build a product, treat teammates well, and articulate the values that drive their company. However, they also have disparate areas of focus and views that balance each other. Drew notes that Arash’s pragmatism helps keep his (Drew’s) natural optimism in perspective. “So we have this kind of yin and yang thing going,” he says. Much press has been given to the challenge early in Dropbox’s existence, when Apple founder and CEO, Steve Jobs offered to buy out Dropbox. Drew notes: “ Nobody knew our business better than we did. And our thinking was, we built something we really loved, and it’s doing well. If the company has this much value today, it's going to be much more valuable in the future.”

In the effort to keep their culture and product fresh, the company holds week-long hacking sessions during which every member works on something, for example, that is particularly interesting or challenging that has not yet been solved. Hackweek, held multiple times a year, has yielded new features and intensive team-building for the entire company. Dropbox has lately been more aggressive with recruiting and acquisitions – a step that Drew and Arash have found very healthy for the company. Drew notes: “We've had a lot of really talented teams join us because overnight we can give them a lot of scale. And, it's a really interesting playground for them.” Notable recent hires include Guido van Rossum, the father of Python, who had previously worked at Google. Arash comments, “At a company like Google where they have over 10k engineers, if you join, you’re not realistically going to be able to spend time with an engineer like Guido. In contrast,” he notes, “at Dropbox with 70 engineers, anyone can eat lunch with Guido.”

Drew’s equanimity also helps keep competition a manageable part of the picture. Drew suggests that more often than not, companies don't reach their potential because of ‘self-inflicted wounds’. “So they hire people that aren't as good,” he says and adds: “…or don't focus on making their users happy or they try to do too many things. And that kind of thing happens a lot more than getting threatened by a competitor. The Dropbox advantage,” Drew explains, “is to focus on making our users really happy – the one problem that we can solve.”

When asked what he would suggest to MIT students who are thinking about working in a startup, Arash suggests that although startups sound risky, “there will always be a job at the larger companies. So take risks early in your career when there is not so much to worry about.” He adds that whether doing a startup or joining an existing one, the number one factor needs to be the quality of people. He says, “I think the best things happen to each of us when we find the most brilliant people and surround ourselves with them. I think it's certainly true in startups and tech where everyone is really smart. Your competitive advantage is being around the absolute best. It just makes such a big difference.”

Nurturing Dropbox Culture and Product

Be sure to tune in for the June 7, 2013 MIT Commencement at which Drew Houston will be speaking. n

At Dropbox, Arash Ferdowsi says he focuses on two things, both very much about quality. He meets everyone that they hire — a measure that allows them to maintain their unique culture. This culture he explains can seem like an extension of

MIT EECS Connector — Spring 2013

59


Alumni Features Sal Khan, SB, MEng ’98

“Yes, it feels utopian to me — living the dream. I get to work with some of the best people in the world literally — putting their talents to what’s important. It all started off with my kind of crazy delusions while sitting in the closet. It's exciting!” Sal Khan became a global superstar when he brought his life experiences and natural inclinations to help others into focus by creating and running the Khan Academy. Since it’s inception in 2008, Khan Academy, as a not-for-profit, has delivered over 240 million online lessons to students worldwide — testament to its mission to provide a free world-class education for anyone, anywhere. Students of all ages have been viewing the over 4000 videos and using the site associated online adaptive exercise platform and online tracking that encourages learning mastery. The story of his journey and his thesis to reinvent education now appears in his book, The One World Schoolhouse, which came out in late 2012. Sal Khan followed his undergraduate education (in math, and electrical engineering and computer science) and his MEng in computer science at MIT in 1998 with several jobs in Silicon Valley for a few years — his first career. Completing his Harvard Business School MBA in 2003, Sal Khan started his second career as a hedge fund manager in Boston. Then came his third (and current) career. Sal Khan has the good fortune to be a natural teacher — as well as a good computer scientist — so he was more than adept at helping his then 7th grade cousin Nadia (based in Louisiana) with a math problem in 2004 while he worked in Boston. Teaching at a distance was easily overcome online and eventually Nadia and her friends and relatives ballooned into a pre-Khan Academy set of viewers of her cousin Sal’s short, entertaining videos ultimately posted on You Tube. Within a few years, Khan’s interests went well beyond hedge fund management. Fortunately, the increasing recognition and support came just in 60

www.eecs.mit.edu

time as Sal reached the tipping point to go on his own. Donors began to surface including Ann Doerr (and her husband John, a Silicon Valley venture capitalist) and then in quick succession, Bill Gates (and his family) who raved about Khan Academy to a large audience at the Aspen Ideas Festival in front of 2,000+ people, support from Google’s Eric Schmidt, and more. By 2009, Sal Khan was fully launched to run his dream — the Khan Academy. Sal Khan got to this point for lots of reasons including his early schooling and his childhood. Although Sal’s mother, a single parent from India, was forced to work extremely hard at many jobs, Sal and his older sister and younger brother grew up in an environment that fostered individuality and curiosity. In his words: “For the most part we had a fairly verbal family which in no small way was a euphemism for argumentative. But you can't discount that. [It] actually makes you good at articulating yourself on the fly.“

Members of the Khan Academy gather for a picture in January, 2012.

Since his experiences growing up in the public schools in New Orleans, Sal Khan has been sensitive to how he was taught. When he was selected to join a class for gifted learners, he says, one of his teachers asked him: “What do you want to do?” His reaction? “I was like “What? You're asking me what do I want to do?” You know, she kind of gave me license to be creative.”

Sal is naturally the kind of person who is not afraid to ask questions to understand something, to get to the heart of a problem or need, to want to teach with passion and effect, and to take on natural leadership roles. (He was president of the Class of 1998 at MIT and president of the student body while a student at Harvard Business School). In so many ways, Sal Khan is what he teaches.

At the heart of his well-formulated thesis about education for humanity, Sal points to the need for allowing each individual (at any age) to proactively seek out learning—a positive and creative act in itself. He proposes that by using the Khan Academy online videos and exercises, which are self-paced and uniquely tracked, anyone (including students in public schools) can reach mastery (100%) of the material without being turned off by lengthy lectures. Teachers can finally give their full passion to individual tutoring while all students can help each other as they progress at different (and similar) rates. His ideas on a self paced, mastery-based model are already being adapted for use in schools in California and elsewhere. (Read more about applying Khan Academy courses and concepts on the Khan Academy website. https://www. khanacademy.org/coach/resources).

"I teach the way I wish I was taught. The lectures are actually coming from me, an actual human being who is fascinated by the world around him."

Sal Khan has been thinking about education for a long time. When he came to MIT, he was keenly aware of how smart students were sometimes struggling while others were not. He

wondered about this. At the same time, he noticed that the pace forced him (and others) to be mercenaries with their time. “I was thinking coldly how to get the most bang for my hour of time,” he recalls. He discovered that he learned best at his own pace — whether getting ‘chunks of understanding’ from a book chapter or working with classmates on a project in the lab. He also realized that the people who understood a subject (such as math, the basis for much of science and engineering) holistically — on an intuitive level — were the ones who learned most deeply. It is interesting that between his junior and senior years as an undergraduate at MIT, Sal Khan was given the Eloranta Fellowship to create education software and he created “Planet Math”. Since then he admits he has lost the domain name and doesn’t know what’s happened to it — though he would like to have it! [We are on it, Sal!] Along the way, Sal also realized that effective online education requires more than just a scripted experience. From his early days teaching his cousin, he knew the value of the human ‘emotive’ delivery. He says, “There should be intonation and cues that you [the teacher] care[s]. A lot of that get's lost when you script things.” Although he misses those early days when he recorded all his videos in a little ‘closet-like’ office, he guards his time to continue preparing and making all the videos for Khan Academy.

MIT EECS Connector — Spring 2013

61


Alumni Features Now the 37+ individuals who work for Sal Khan are applying these concepts to the growing Khan Academy. They come from all over (including EECS and MIT) and are highly gifted computer scientists and engineers who are excited to contribute to this effort. While, the focus is to effectively deliver online mastery learning experiences in all the major core topics such as physics, chemistry and biology, much is happening in the real-time teaching and learning potential offered through Khan Academy. Last summer Sal and members of the Academy ran a camp designed to discover what was possible in the physical learning environment interfacing with young students — a project simulation-based experience. He notes, “There is a part of our DNA that really wants to interface with real kids to see what and how they can learn.” Well over a thousand applications were made for the class of 60 — another indication of the growing interest in the Khan Academy approach. Is there competition? Sal Khan says, yes, but not in the traditional sense. He explains: “There are others starting to give tools — some for profit, some not — like Massive Open Online Courses (MOOCs). I don’t know whether you call them competition. They are in the same space. We want to be on the cutting edge. So, if they’re doing something cool, we want to be able to leverage that. It’s kind of like the competition you have with friends. You want to do well but you want to continue the friendship.” In terms of MIT’s digital learning tools, Sal Khan was extremely impressed when Open Courseware was announced in 2001. His reaction to MITx was a lot more personal. He notes, “Later [in late 2010], when they set up MITx, they referred to ‘KSVs’ [Khan Style Videos]. I was blown away. Now they’re citing me!” (He was and is a big supporter of edX/MITx.) In early 2012, the MIT School of Engineering launched a series of videos produced by MIT students and aimed at K-12 learners. Sal Khan helped with this project and the MIT+K-12 videos are integrated on the Khan Academy website.

Top: Sal Khan works with 6th grade students and teachers at Eastside Prep in East Palo Alto, California. Bottom: Sal Khan spends time ‘in the field’ checking on students using Khan Academy materials in the Egan, California classroom.

And, Sal Khan’s life observations about the holistic learning experience — where ideally, you learn by connections between content — have framed his approach. He notes, “When math or science was taught, it was often not in the most emotive way — often coming from one unit to another. It was very disjoint; connections weren’t drawn across subjects. Even within a subject connections were often lost.”

62

www.eecs.mit.edu

This may have been about the same time that Sal Khan was approached to speak at MIT. He reflects: “Speaking at graduation was like a Nobel Prize, literally. If someone had told me while I was at MIT that I would be the commencement speaker, I would have said "If that happens to me the day before I am 80 years old, that's awesome.” Among the many inspirational personalities Sal Khan followed as a child (including his older sister), late night comedian Johnny Carson ranks high. He says, “Literally from age 3 until 11 when he retired, I was a regular watcher. I thought there was no more fun thing to do than to watch. Clearly I had no bed time!” But he admits he didn’t get most of the jokes. But, he did get the idea of entertaining — for education’s sake.

Susie Wee, ’90, SM ’91, PhD ’96

“A big theme in my life has been teamwork and collaboration — helping people work together in teams to achieve bigger things than they can do themselves. Those are the things that I do naturally.”

Several months ago Susie Wee (’90, SM ’91, PhD ’96) took on a new role at Cisco as the VP and Chief Technology Officer of Networked Experiences. As with other transitions in her life and career, her abilities as a strong team builder and a well-grounded technologist have propelled her into new roles and challenges. Although she grew up in a small town in western New York — not a Silicon Valley or a Cambridge hub — Susie Wee was intensely interested in computers and programming. In fact, she not only went to programming classes that started at 6 am, but she stayed up all night programming on the family’s Apple IIe computer. When Susie applied to MIT and received her acceptance letter, her father, who is a medical doctor (who really wanted to be an engineer), brought the letter to her in the middle of her high school class. She recalls, “He was so happy it was like he got into MIT!” Her older brother also went to MIT and is a medical doctor as well. As a Course VI student, Susie especially loved the double 0’s — 6.001, 6.002, etc. In particular, she loved 6.003 for which Prof. Hal Abelson was her recitation instructor. She says, “I loved the way he taught 6.003 — working with Fourier transforms and signals and systems gave me a great way to think about things and got me interested in optics and image processing.” As she seized opportunities throughout her undergraduate years and into her masters at MIT, Susie Wee built on her love

for signal processing. Taking advantage of an Institute-wide internship program, she worked summers at the Jet Propulsion Laboratory, ultimately completing her masters degree and building her knowledge of optics. And, working as a UROP in the Media Lab with Prof. Ted Adelson, Susie was able to extend her knowledge of 6.003 and her growing interest in optics by doing image and video processing.

A formative graduate school experience Continuing in optics and image processing at MIT EECS, Susie Wee joined the lab of the late Prof. Bill Schreiber as his last PhD student and later she found out she was his first female PhD graduate. She notes that it was a great experience working with Prof. Schreiber, who “always just treated me like all the others.” She particularly appreciated the fact that Prof. Schreiber had returned to his lab from retirement to continue work that aimed at practical application. She notes: “I loved that he was not only advancing theory but he was also making leading edge practical systems. Over his career he made black and white television into color television, black and white printing into color printing, and standard definition television into high definition television. He was heavily involved with industry and with government policy, making sure that things went in the right direction, really looking at the broader issues at play.” Susie Wee and Mike Polley (’89, SM ’90, PhD ‘96) were Prof. Schreiber’s last two PhD students, and they studied closely with their now spouses John Apostolopoulos (’89, SM 91,

MIT EECS Connector — Spring 2013

63


Alumni Features PhD ’96) and Rajni Aggarwal (’89, SM ’90, PhD ‘96). Susie and Mike worked together on the technology aspects of a joint project on a new approach to building a high-definition television (HDTV) system. Susie designed the source code and Mike designed the channel coder. John worked on the MIT Grand Alliance HDTV system for Prof. Jae Lim. It was impressive to her that many of Prof. Schreiber’s earlier students were designing and building the prototype HDTV systems in different companies — before the standards were established. She also liked the fact that because Prof. Schreiber was retired, she was sharing the lab with Professors Jae Lim and Al Oppenheim and their students. “It was really great to have so many colleagues to know and work with,” she notes.

Naturally drawn to Teams and Technology Always an avid ice hockey player, Susie Wee found that her love not only for the sport but also for being part of a team eventually resulted in her becoming the ice hockey captain and ultimately a team coach during her 10 years at MIT. In fact, when she accepted the Women in Technology International (WITI) Hall of Fame award in 2010, she cites and compares her experiences in both playing and coaching ice hockey to being a woman in technology, where creating a winning move or strategy can make the difference. When she interviewed at Hewlett Packard Labs in Silicon Valley in 1996, Susie Wee gave a talk. She figured that it was standard that not only was this advertised but also the room was packed. It was affirming that her presentation, which was based on her thesis work, was considered unique and cutting edge by the HP researchers. She says, “When I got there and started working I found that not all interview talks are so well attended. I also found that technically, I was really prepared.” Several other factors were striking to her. As one of very few females in the HP Labs, she says: “I didn’t look like anybody else. Most were older men with button down shirts and khakis.” Nevertheless, she and her work were valued and she brought a new dimension to what they wanted to achieve. She reasoned: “The way I am and I think most MIT people are, I could develop technology all day long. At that stage I had no trouble producing things, being relevant and pushing the state of the art forward.” Two years into her work, when she had a performance review with her boss, she noted the contrast between her workweek and her team-sports-filled weekends. HP then recognized Susie Wee’s potential, and she went from supervising an intern to leading projects and managing teams. Ultimately Susie Wee was leading big projects with global, cross-functional teams and leading collaborations with companies such as NTT DoCoMo working on mobile video systems for 3G and 4G 64

www.eecs.mit.edu

wireless networks and incubating new product areas such as the HP Halo immersive video conferencing system and the HP OpenCall Media Platform. At this time, Susie also led a collaboration with EECS faculty at MIT including Professors Anantha Chandrakasan, Muriel Médard and Greg Wornell to outline a plan for an MIT Wireless Center as well as establish an HP-MIT Researchers in Residence program. Under this program, which ran from 2000 for nearly ten years, MIT students, particularly those in between their masters and PhD, came out to work in the HP Labs during IAP, creating a formative experience toward their PhD thesis work. Ten years into her HP Labs experience, Susie Wee moved into the business — first into the PC group, where she was the Vice President of the Experience Software Business, where she ran software development for HP’s PCs and where she took her first CTO role. She notes about this point when she was stepping out of research, “You could say that HP Labs is like a natural extension of MIT, doing graduate school and research. Then basically I went up the chain in HP Labs and became a lab director. But then I took the career step to move into the business—this was something completely different for me.”

Transitioning is like starting over — both challenging and exciting “Becoming a vice president in charge of running a software business,” Susie Wee notes, “required learning the current way the business does things and then learning how I could make an impact on the business with my different perspective and experience.” In her new role as CTO of Client Cloud Services, in HP's Personal Systems Group, Susie ultimately worked on HP Halo, which involved developing face-to-face telepresence high-end video conferencing systems. In a Tedx Bay Area Women talk in 2010, she described this kind of helping people to communicate as “Making Local Global and Global Local” — something, which she obviously loved and believed in.

“The neat thing about the MIT experience is that you are hanging out with really smart people — the people you live with, the people you hang out with and grow with. You take it all for granted at the time, but when you leave MIT, you realize that you have pretty much worked with the smartest people you are ever going to bump into." Now, Susie Wee is working across all of Cisco’s products. With the networking and data center teams she looks at software and application-centric networking both for the enterprise as well as for service providers and data centers. Her efforts aim to bring an experience centric approach to the area — understanding what the operational experience is for these networks and converged infrastructure and how it can be improved with the fundamental shift in architecture towards software-defined networks. She notes: “The area of software and application-centric networking is being formed. It has the chance of being defined poorly or defined well. If it gets defined well then there are tremendous opportunities. It is important to shift from a technology only approach to an experience- and technology-driven approach.”

grow with. You take it all for granted at the time, but when you leave MIT you realize that you have pretty much worked with the smartest people you are ever going to bump into. This raises your game.” She has found this helps her as she raises the bar in her career advancements.

Throughout Susie Wee’s career moves since MIT, she has kept several constants to guide her. One is her experience living with people at MIT. She says: “The neat thing about the MIT experience is that you are hanging out with really smart people — the people you live with, the people you hang out with and

* In Remembrance of Professor William F. Schreiber, Nov. 21, 2009 entry in Susie Wee’s blog. n

At the same time, she remembers her PhD advisor Prof. “Bill” Schreiber, for whom she spoke at the celebration of his life held at the MIT Faculty Club on November 21, 2009. She spoke about ‘Schreiber-isms’ — lessons, which he left with all who knew him—particularly his students. Among these, “Bill advised me not to be an analyst, but to build things that add value to the world.” It appears to be a lesson that Susie Wee has taken to heart.

Although Susie admits that she “tends to stick around”—as she had for 10 years at MIT and 15 years at HP—she accepted an offer to work as CTEO (Chief Technology and Experience Officer) of Collaboration at Cisco. Right away, she was encouraged to build the position to combine technology and experience. She notes: “I was able to work across all of the different businesses: video conferencing, web conferencing, unified communication, and instant messaging and presence —mixing these different areas into an integrated collaboration experience. This was in April 2011, when Susie moved up to Vice President and Chief Technology and Experience Officer, CTEO of Collaboration.

MIT EECS Connector — Spring 2013

65


Donor Recognition and Appreciation From the Department of Electrical Engineering and Computer Science at MIT, we extend our thanks to the generous donors listed below who made gifts to the Department this past fiscal year 2012 (July 1, 2011-June 30, 2012). We have attempted to list all donors of $100 or more to EECS in this time period unless anonymity was requested. Although care has been taken in the preparation of this list, some errors or omissions may have occurred; for these we extend our sincere apologies. If you designated your gift to the EECS Department and your name does not appear here or is incorrectly listed, please bring the error to our attention. All donor recognition categories are exclusive of corporate matching gifts. *deceased

BENEFACTORS $100,000 – 999,999 Thomas Kailath SM ’59, ScD ‘61 Alan L. McWhorter ScD ‘55 Franklin Quick Jr. ‘69, SM ‘70

PATRONS $50,000 – 99,999 David R. Fett ’77, SM ‘77 Daniel B. Grunberg ’82, SM’83, PhD ‘86 James K. Roberge ’60, SM ’62, ScD ‘66 Raymie Stata ’90, SM ’92, ScD ‘96

SPONSORS $10,000 – 49,999 James D. Ahlgren ‘55 Weng C. Chew ’76, EE ’78, SM ’78, PhD ‘80 Sunlin Chou ’66, SM ’67, EE ‘68 Paul R. Drouilhet, Jr ’54, SM ’55, EE ‘57 Renée Finn Steven G. Finn ’68, SM ’69 Paul A. Green II ‘73 John V. Guttag Olga P. Guttag Michael G. Hluchyj ’79, SM ’79, PhD ‘82 William W. Irving, Jr ’87, SM ‘91 EE '92 MEng. '92, PhD'95 Charlene C. Kabcenell ‘79 Dirk A. Kabcenell ‘75 Kurt A. Locher ’88, SM ‘89 Barry Margolin ‘83 Chew C. Phua PhD ‘81 Nils R. Sandell SM’71, EE ’73, PhD ‘74 Richard L. Townsend SM ’59, EE ‘60

SUPPORTING MEMBERS $5,000 – 9,999 Arthur A. Gleckler ’88, SM ‘92 John C. Hardwick ’86, SM ’88, PhD ‘92 Charlene C. Kabcenell ‘79 Dirk A. Kabcenell ‘75 Ronald B. Koo ’89, SM ‘90 Alexander Kusko SM ’44, ScD ‘51 Kenneth W. Nill ’61, SM ’63, PhD ‘66 Mary M. Rollins Malcom L. Schoenberg ‘45 Andrew F. Stark ’97, MEng ‘98 Gunter Stein John D. Summers SM ‘84

66

Edward G. Tiedemann PhD ‘87 Alan S. Willsky ’69, PhD ‘73 Ronald E. Zelazo ’66, SM ’67, EE ’69, PhD ‘71

SUSTAINING MEMBERS $1,000 – 4,999 Donald J. Aoki SM ‘79 Terrance R. Bourk SM ’70, PhD ‘76 Jose M. Brito Infante SM ’62, EE ‘63 Charles G. Bures ‘69 Arthur C. Chen ’61, SM ’62, PhD ‘66 David R. Cheng ’04, MEng ‘05 Richard J. Codding SM ‘66 Fernando J. Corbató PhD ‘56 Boris D. Cukalovic ’05, MEng ‘06 Donald A. Dobson SM ‘51 David H. Doo ‘77 Robert R. Everett SM ‘43 Robert M. Fano ’41, ScD ‘47 Jenny M. Ford ’81, SM ‘82 G. David Forney SM ’63, ScD ‘65 Janet A. Fraser SM ‘84 Robert W. Freund EE ’74, SM ‘74 Edward C. Giaimo ’74, SM ‘75 Sol W. Gully Joshua Y. Hayase ’52, EE ’57, SM ‘57 Jerrold A. Heller SM ’64, PhD ‘67 Steven J. Henry ’72 SM ‘73 Gim P. Hom ’71, SM ’72, EE ’73, SM ‘73 Tareq I. Hoque ’88, SM ‘88, SM ‘92 William F. Kelly SM ’61, EE ‘63 Byungsub Kim SM ’04, PhD ‘10 Richard Y. Kim ’83, SM ‘88 Ernest R. Kretzmer SM ’46, ScD ‘49 Yang-Pal Lee ‘72 Frederick J. Leonberger SM ’71, EE ’72, PhD ‘75 Anthony J. Ley SM ‘63 Frank J. Liu EE ‘66 John I. Makhoul PhD ‘70 Henrique S. Malvar PhD ‘86 Terrence P. McGarty Jr SM ’66, EE ’69, PhD ‘71 Sharon E. Perl SM ’88, PhD ‘92 Wendy Peikes ‘76 Paul L. Penfield, Jr ScD ‘60 Robert J. Petrokubi SM ‘68 Lisa A. Pickelsimer SM ‘92 Marina E. Pocaterra Silva SM ‘87 Clark J. Reese SM ’69, EE ‘70 Ellen E. Reintjes ‘73 Roger A. Roach Joseph J. Rocchio ’57, SM ‘58

MIT EECS Connector — Spring 2013

Melanie B. Rudoy SM ’06, PhD ‘09 William L. Sammons ’43, SM ‘44 John E. Savage ’61, SM ’62, PhD ‘65 Richard J. Schwartz SM ’59, ScD ‘62 Charles L. Seitz ’65, SM ’67, PhD ‘71 Carol Tucker Seward ‘47 Paul J. Shaver SM ’62, ScD ‘65 Burton J. Smith SM ’68, EE ’69, ScD ‘72 Dorothy D. Smith EE ‘72 David Smullin David L. Sulman SM ‘69 Karl Sun ’92, SM ’93, SM ‘97 Donald L. Tatzin ’73, ‘75 Aurelie Thiele SM ’00, PhD ‘04 Richard D. Thomton SM ’54, ScD ‘57 Joseph E. Wall EE ’76, SM ’76, PhD ‘78 David Wang ’00, MEng ‘00 Harold M. Wilensky ‘70 Joseph F. Wrinn ‘75 Anthony Yen SM ’87, EE ’88, MEng ’88, PhD ’92 Dale A. Zeskind EE ’76, SM ‘76 Bazil R. Zingali ‘58 $500 – 999 Abeer A. Alwan SM ’86, EE ‘87, MEng ‘87, PhD ‘92 Bill W. Agudelo SM ‘82 Chalee Asavathiratham ’95, MEng ’96, PhD ‘01 Murat Azizoglu PhD ‘91 Richard A. Barnes ‘68 Robert V. Baron ’71, EE ’77, SM ‘77 Manish Bhardwaj SM ’01, PhD ‘09 Geoffrey F. Burns SM ’89, PhD ‘92 Woo S. Chang SM ’99, PhD ‘03 Shu-Wie F. Chen ‘86 Douglas J. Deangelis SM ‘06 Peter V. Dolan SM ‘79 Matthew L. Fichtenbaum ‘66, SM ‘67, EE ‘68 Michael D. Gerstenberger EE ’85, SM ‘85 Stephen D. Hester ‘63 Caroline B Huang SM ’85, PhD ‘91 David F. Huynh SM ’03, PhD ‘07 David L. Isaman SM ’70, PhD ‘79 Bartley C. Johnson EE ’81, SM ’81, PhD ‘86 Matthew J. Johnson SM ‘10 John S. Keen ScD ‘94 Wayne G. Kellner SM ’57, ScD ‘63 Shing Kong ‘94 James W. Lambert ‘76 Kristina S. Lambert EE ’76 Stan Y. Liao ’91, SM ’92, PhD ‘96 Nathan A. Liskov ‘60

Herschel H. Loomis Jr PhD ‘63 Charles I. Malme SM '58, EE '59 Warren A. Montgomery EE ’76, SM ’76, PhD ‘79 Robert C. Moore ’70, SM ’76, PhD ‘79 Joel Moses PhD ‘67 Phillip T. Nee SM ’94, PhD ‘99 Carl E. Nielsen Jr SM ‘58 Paola F. Nisonger SM ‘79 Robert L. Nisonger SM ‘78 Cynthia A. Phillips SM '85, PhD '79 Alexander L. Pugh SM ’53, EE ‘59 James D. Roberge ‘84 Murray A. Ruben EE ’64 SM ‘64 Nora L. Ryan ‘84 Edward M. Singel EE ’75, SM ‘75 Christopher E. Strangio EE ’76, SM ‘76 Joan M. Sulecki SM ‘83 Michael L. Telson ’76, SM ’69 EE ’70, PhD ’73 James M. Thompson ‘77 Oleh J. Tretiak SM ’60, ScD ‘63 John C. Ufford SM ‘75 Junfeng Wang SM ’97, PhD ‘99 Timothy A. Wilson ’85, SM ’87, ScD ‘94 Robert A. Young PhD ‘68 $250 – 499 Robert L. Adams SM ‘69 Roger K. Alexander SM ‘91 Tao D. Alter SM ’92, PhD ‘95 Michael S. Basca SM ‘00 H. E. Blanton SM ’49, EE ‘55 Toby Bloom SM ’79, PhD ‘83 Richard W. Boberg SM ‘73 Michael S. Branicky ScD ‘95 Randal E. Bryant SM ’77, EE ’78, PhD ‘81 Matthew S. Burnside ’00, MEng ’01, MEng ‘02 Julian J. Bussgang ‘51 Charles H. Campling SM ‘48 John Y. Cha ’91 SM ‘92 David L. Chaiken SM ’90, PhD ‘94 Ga-Yin L. Chan SM ‘87 Daniel K. Chang SM ‘92 Brian A. Chen SM ’96, PhD ‘00 Harry H. Chen SM ‘76 Rene G. Cornejo ‘84 Charles H. Cox ScD ‘79 Joel S. Douglas SM ’91, PhD ‘95 Arthur Evans Bryan A. Ford SM ’02, PhD ‘08 Takeo Fukuda SM ‘74 Kenneth W. Golf SM ’52, ScD ‘54 Nicholas Gothard SM ‘62 Richard S. Grinnell ’92, SM ‘93 Allan R. Gunion SM ‘60 Gudmundur Hafsteinsson Walter C. Hamscher SM '83, PhD '88 Stephen M. Hannon SM ’87, EE ’88, MEng ’88, PhD ‘90 Steven G. Herbst ’10, MEng ‘11 John S. Hill SM ‘60 Jerry L. Holsinger PhD ‘65 Jeffrey Jouppi S ’95 EE Steven Kamerman ‘73

Stephen T. Kent SM ’76, EE ’78, MEng ’78, PhD ‘81 Zafar M. Khan ’79, SM ‘85 Thomas F. Klimek SM ‘59 Wolf Kohn SM ’74, PhD ‘78 Stephen F. Krasner ’71, SM ‘91 Yu-Ting Kuo SM ‘94 David B. Leeson SM ‘59 Ying Li SM ’89, EE ’93, MEng ’93, PhD ‘94 Chia-Liang Lin PhD ‘95 Francis C. Lowell Jr SM ’64, EE ‘65 Glendon P. Marston ScD ‘71 Emin Martinian SM ’00, PhD ‘04 Christopher S. McNulty ’98, MEng ‘99 Scott E. Meninger SM ’99, PhD ‘05 Keith S. Nabors SM ’90, PhD ‘93 Stephen D. Patek SM ’94, PhD ‘97 Vinay Pulim ’97, MEng ’99, PhD ‘08 Thomas J. Richardson PhD ‘90 Larry S. Rosenstein ’79, SM ‘82 Michael G. Safonov ’70, SM ’71, EE ’76, PhD ‘77 Paul S. Schluter EE ’76, SM ’76, PhD ‘81 Howard Schneider ‘79 Jean-Pierre Schott EE ’82, MEng ’82, SM ’82, PhD ‘89 Sarah E. Schott ‘83 Philip E. Serafim SM ���60, ScD ‘64 Howard J. Siegel ‘71 Guy M. Snodgrass SM ‘00 Birgit L. Sorgenfrei SM ‘93 David A. Spencer SM ’71, EE ‘72 Steven V. Sperry SM ‘78 Daniel D. Stancil SM ’78, EE ’79, PhD ‘81 David L. Standly SM '86, PhD '91 Russell L. Steinweg ‘79 Charles A. Stutt ScD ‘51 Olivia Tsai ‘03 James N. Walpole SM ’62, PhD ‘66 Kathleen E. Wage SM ’94, MEng ’96, PhD ‘00 Eric S. Wang ‘09 J. S. Wiley ‘73 John A. Wilkens PhD ‘77 Lucile S. Wilkens PhD ‘77 Harvey M. Wolfson EE’74, SM ‘74

$100 – 249 Jeffrey D. Abramowitz SM ‘85 Boon S. Ang SM ’93, MEng ’98, PhD ‘99 Mark Asdoorian ’98, MEng ‘98 Michael A. Ashburn SM ‘96 Eileen J. Baird SM ‘87 David R. Barbour SM ‘61 Robert L. Baughan Jr SM ‘49 Arlyn W. Boekelheide SM ‘52 Gary C. Borchardt PhD ‘92 Hardy M. Borland SM ‘57 Paul D. Bosco John R. Brennand SM ‘59 Tom P. Broekaert SM ’89, PhD ‘92 Philippe Brou SM ’81, PhD ‘83 Gretchen P. Brown EE ’74, SM ‘74 John F. Buford ’79, SM ‘81 Wayne M. Cardoza SM ‘71

Christopher Carl ‘60 John C. Carter SM ‘78 Philippe M. Cassereau EE ’85, SM ‘85 Valentino E. Castellani SM ‘66 Upanishad K. Chakrabarti ‘94 David Chase SM ’64, PhD ‘67 Kan Chen SM ’51, ScD ‘54 Seong Hwan Cho MS ’97, SM ’97, PhD ‘02 Myung Jin Choi SM ’07, PhD ‘11 Nelson C. Chu SM ‘90 Douglas R. Cobb SM ‘65 Jack D. Cowan SM ‘60 Leigh C. Cropper ‘69 David R. Cuddy EE ’74, SM ‘74 Susan R. Curtis SM ’82, PhD ‘85 Stefano I. D’Aquino SM ‘91 Bahman Daryanian ’77, SM ’80, SM ’86, PhD ‘89 George A. Davidson SM ‘56 Wilbur R. De Hart* SM ’46, EE ‘51 Douglas R. Denison PhD ‘99 Lyric P. Doshi ’08, MEng ‘10 Jon Doyle SM ‘77, PhD ‘80 Adam M. Eames ’04, MEng ‘05 Carol Y. Espy-Wilson SM ’81, EE ’84, MEng ’84, PhD ‘87 Kenneth W. Exworthy SM ‘59 Maya S. Farhoud SM ’97, PhD ‘01 James G. Fiorenza PhD ‘02 Richard E. Fitts SM ’55, PhD ‘66 Paul J. Fox EE’73, SM ‘73 Roy E. Fraser SM ‘64 Thomas H. Gauss SM ‘73 Diane C. Gaylor SM ‘87, PhD ‘89 Michael A. Gennert ‘80, SM ‘80, ScD ‘87 Lewis K. Glanville* SM ‘62 William G. Glass SM ‘50 Thomas J. Goblick Jr SM ’58, EE ’60, PhD ‘63 David J. Goldstone ‘89, SM ‘91 Aaron A. Goodisman ’90, SM ‘91 Randall V. Gressang SM ’66, EE ‘67 Stephen E. Grodzinsky ’65, SM ‘67 Kush Gulati PhD ‘01 David H. Guttag ‘05 Wayne H. Hagman SM ‘81 William A. Harrison EE ’84, SM ‘84 John M. Heinz SM ’58, EE ’59, ScD ‘62 Wendi B. Heinzelman SM ’97, PhD ‘00 Walter O. Henry SM ‘52 Herbert L. Hess SM ‘82 Melanie E. Holland ‘90 Roger A. Holmes SM ‘58 Merit Y. Hong ’84, SM ’87, PhD ‘91 Berthold K. Horn SM ’68, PhD ‘70 Charles A. Hornig ‘79 Syed Z. Hosain ‘79 Paul K. Houpt PhD ‘75 Edward P. Hsieh SM ‘64 Caleb W. Hug SM ’06, PhD ‘09 Hing-Loi A. Hung ‘68 Joseph A. Hunt SM ‘63 Ernest G. Hurst Jr ’60, SM ’65, PhD ‘67 John D. Jackman ’06, MEng ‘07 Stephen C. Jens SM ‘83 Hsin-Kuo Kan SM ’73, ScD ‘77

MIT EECS Connector — Spring 2013

67


Donor Recognition and Appreciation Karrie Karahalios ’94, MEng ’95, SM ’98, PhD ‘04 Richard L. Kautz SM ’72, PhD ‘75 Kenneth P. Kimi Jr SM ‘81 Richard D. Klafter ‘58 Gregory A. Klandeman SM ‘95 Ernest S. Kuh SM ‘50 John L. Kulp Jr ’71, SM ’73, EE ’74, PhD ‘78 Shawn Kuo SM ‘04 Mark Kuznetsov EE ’82, SM ’82, ScD ‘86 Shang-Chien Kwei ‘05 John D. Lattanzi SM ’79 Christopher T. Lee SM ’62, EE ‘66 Jay K. Lee SM ’81, EE ’82, PhD ‘85 Michael Lee SM ‘95 Benjamin J. Leon SM ’57, ScD ‘59 Alan Levin ‘72 Alexander H. Levis ’63, SM ’65, MEng ’67, ScD ‘68 Kevin A. Lew SM ‘95 Wendy Liu-Battaiora ’88, SM ‘89 Sylvie J. Loday SM ‘01 John M. Ludutsky Jr ‘64 Allen W. Luniewski EE ’77, SM ’77, PhD ‘80 William F. Maher Jr SM ‘80 John A. Mallick ’73, EE ’76, SM ’76, ScD ‘79 Carla Marceau SM ‘70 Steven I. Marcus SM ’72, PhD ‘75 Elizabeth A. Marley SM ’96, PhD ‘00 Uttara P. Marti ’03, MEng ‘05 Patrick J. McCleer SM ‘72 John C. Mitchell SM ’82, PhD ‘84 Gerardo H. Molina SM ‘86 Lajos Molnar ’97, MEng ‘98 Guy E. Mongold Jr SM ‘59 Paul Moroney ’74, EE ’77, SM ’77, PhD ‘79 Ivan A. Nausieda PhD ‘09 Osama M. Nayfeh SM ’04, PhD ‘09 Stephen D. Offsey ‘86 Carmelo A. Palumbo SM ‘02 Hugh M. Pearce SM ’66, EE ‘67 David J. Perreault SM ’91. PhD ‘97 Thomas J. Perrone ‘65 Marvin E. Petersen SM ‘57 Brian K. Pheiffer SM ‘92 Michael Piech ‘90 Damian O. Plummer ‘02 Charles C. Post ’09, MEng ‘11 Matthew H. Power ’85, SM ’88, PhD ‘93 Edward J. Powers Jr SM ‘59 Aditya Prabhakar ’00, MEng ‘01 James C. Preisig EE ’88, MEng ’88, SM ’88, PhD ‘92 Robert A. Price SM ‘53 Steven D. Pudar SM ‘92 James F. Queenan ‘63 Nadir E. Rahman ’95, MEng ‘96 Richard H. Rearwin SM ‘54 Louis R. Records SM ‘78 John A. Redding SM ‘76 Howard C. Reeve SM ‘83 Barry D. Rein ‘60 Mark B. Reinhold SM ’87, PhD ‘93 John F. Reintjes Jr ‘66 Chester D. Reis Jr SM ‘70 Thomas D. Rikert ’99, MEng ‘99

68

Gail C. Ruby SM ‘81 William D. Rummler SM ’60, EE ’61, ScD ‘63 Karen B. Sarachik SM ’89, PhD ‘94 Sunil K. Sarin SM ’77, EE ’78, PhD ‘84 Frank M. Sauk ’74, SM ‘77 Amrita Sawhney Terrence L. Saxton SM ‘66 Ronald W. Schafer PhD ‘68 Roger R. Schell PhD ‘71 Joel E. Schindall ’63, SM ’64, PhD ‘67 Jeremy H. Scholz ‘05 Russell S. Schwartz ’96, MEng ’96, PhD ‘00 David A. Segal ‘89 Orit A. Shamir ’06, MEng ‘08, PhD ‘12 Edmund M. Sheppard SM ‘58 Donald L. Snyder SM ’63, PhD ‘66 Gary H. Sockut SM ‘74 Shunmugavelu D. Sokka SM ’99, PhD ‘04 Avron N. Spector ’54, SM ‘57 John M. Spinelli SM ’85, PhD ‘89 Clifford S. Stein SM ’89, PhD ‘92 Eric H. Stern ‘73 Andrew C. Sutherland ’02, MEng ‘03 Corina E. Tanasa SM ‘02 Kah Keng Tay ‘08, SM ‘08 Susak Thongthammachat SM ‘67 Kimball D. Thurston ‘94 David A. Torrey SM ’85, EE ’86, PhD ‘88 Charles D. Trawick SM ‘80 Constantine N. Tziligakis SM ’96, SM ‘99 Olga Y. Veselova SM ‘03 Danh T. Vo ’08, MEng ‘09 Edward F. Walker SM ‘84, EE ;‘85 Lawrence C. Wang ’99, ’00, MEng ‘03 Susan S. Wang ‘83 Xiaohong Wang Jennifer Welch SM ’84, PhD ‘88 Daniel S. Weld SM ’84, PhD ‘88 Robert J. Wenzel EE ’74, SM ‘74 Gary L. Westerlund David F. Winchell SM ‘71 James F. Womac SM ’66, PhD ‘72 Ying-Ching E. Yang SM ’85, EE ’86, MEng ’86, PhD ‘89 Roy D. Yates SM ’86, PhD ‘90 Robert D. Yingling SM ‘68 Ryan E. Young ‘08 Marcus Zahn ’67, SM ’68, EE ’69, ScD ‘70 Lingling Zhang Wei Zhu

MIT EECS Connector — Spring 2013

Remember This?

FOUNDATIONS AND CORPORATIONS Adobe Systems Incorporated ARM Inc Andrea Bocelli Foundation Barracuda Networks Inc Boeing Company Broadcom Foundation Cognex Corporation Cypress Semiconductor Digital Voice Systems Inc Dropbox Inc Exxon Mobil Corporation Getco LLC Google Inc Intel Corporation Lockheed Martin Corporation Northrop Grumman Corporation Oracle America Inc Stigall Consulting Group Texas Instruments Inc TriniTEQ Inc Two Sigma Investments LLC Verizon Yahoo!

1

2 Do any of these images look vaguely or very familiar? We invite you (our alumni) to share your caption and memories. Please identify yourself with your year of graduation/degree including the lab and/ or faculty you worked with, and provide the photo number, that you are identifying. We will publish the captions below each photo in the alumni section of our website. To enter your feedback use the form on the EECS website at: eecs.mit.edu/people/alumni. Click on the link to: 2013 Remember This?

3

4


NON PROFIT ORG. U.S. POSTAGE

PAID Massachusetts Institute of Technology 77 Massachusetts Avenue, Room 38-401 Cambridge, MA 02139-4307

CAMBRIDGE, MA PERMIT NO. 54016

You may make corrections to your name, mailing address, or email address by sending email to mitalum@mit.edu or calling Alumni Records at 617-253-8270 or 800-648-1865

Goldwasser and Micali: Recipients of the ACM 2012 Turing Award MIT EECS professors Shafi Goldwasser and Silvio Micali have won the Association for Computing Machinery’s (ACM) A.M. Turing Award for their pioneering work in the fields of cryptography and complexity theory. The two developed new mechanisms for how information is encrypted and secured, work that is widely applicable today in communications protocols, Internet transactions and cloud computing. They also made fundamental advances in the theory of computational complexity, an area that focuses on classifying computational problems according to their inherent difficulty. Goldwasser and Micali were credited for “revolutionizing the science of cryptography” and developing the gold standard for enabling secure Internet transactions. The Turing Award, which is presented annually by the ACM, is often described as the “Nobel Prize in computing” Read more on the EECS website: www.eecs.mit.edu Silvio Micali and Shafi Goldwasser, winners of the A.M. Turing Award. [Photo: Jason Dorfman CSAIL/MIT]


the 2013 MIT EECS Connector