2007
September 2007
Dear Computer Science Alumni and Friends,
On behalf of the Department of Computer Science, it is my pleasure to extend our warmest greetings and share with you some of the department’s major news from the 2006-07 academic year. Highlights in this newsletter include the arrival of new faculty, promotions and retirements of existing faculty, a brief glimpse at research activities, and the celebration of awards and honors bestowed upon department members.
Our department continues to be ranked at the top of all computer science departments nationally and internationally. Our faculty are deeply involved in teaching and conduct research on a broad range of topics at the forefront of computing. We continue to attract the best and brightest students from around the world.
New Faculty
I am pleased to welcome Mehran Sahami back to the department. Mehran rejoins us as associate professor (teaching) and associate chair for undergraduate education. Mehran was a top-rated lecturer in the department from 2001 to 2006. He also has worked at E.piphany and in the research group at Google. His PhD work with Daphne Koller applied machine learning methods to information retrieval problems. Mehran will be leading a revision of our undergraduate curriculum (more on this below).
Promotions
The department had three faculty promotions this year. Sebastian Thrun was promoted to the rank of full professor. Ron Fedkiw and David Mazières were both promoted to the rank of associate professor with tenure.
Sebastian Thrun is best known for leading Stanford’s victorious effort in the 2005 DARPA Grand Challenge, the robotic vehicle race across the Mojave Desert. Sebastian’s area of research is the processing of information in physical spaces under uncertainty, predominantly in the context of robotics. He is a pioneer in applying probabilistic methods to robotics and is considered a major leader in mobile robotics. He is particularly well known for his work on simultaneous localization and mapping. Sebastian’s results of applying Bayesian and Markov estimation and expectation maximization have revolutionized the field. Sebastian is unique in that he both develops novel theoretical techniques and applies them to real-world robotic systems. Sebastian is currently working on a robot car named Junior, Stanford’s entry for the DARPA Urban Challenge, which will take place this November.
Ron Fedkiw’s research focuses on numerical algorithms that enable researchers to represent real- world phenomena on the computer. His research is a synergy between computational physics, which focuses on the accurate simulation of actual phenomena, and physics-based computer graphics, which creates visual images that are appealing and plausible. Recently, Ron has taken on the problem of musculoskeletal simulation.
Ron has contributed animation effects to many well-known feature films. He is the recipient of the 2005 Association for Computing Machinery SIGGRAPH Significant New Researcher Award for his contributions to the field of computational fluid dynamics, including his work on the simulation of multi-material and highly reactive flows. He also received the 2005 National Academy of Sciences Award for Initiatives in Research “for his many innovations in the modeling and numerical simulation of flows, and his pioneering contributions to physics-based computer graphics.”
David Mazières joined us from New York University two years ago and conducts research in systems security. David’s work includes the Asbestos and HiStar operating systems. Asbestos allows system designers to build secure websites using decentralized information flow control and to obtain much stronger security properties than on traditional operating systems such as Unix or Windows. HiStar builds on the Asbestos system and addresses a number of hard problems related to covert channels and resource allocation that were not solved in Asbestos.
David’s previous contributions include SFS, a secure Internet file system that provides a wide range of key management systems controlled by the user to allow file sharing over untrusted networks; SUNDR, a network file system designed to store data securely on untrusted servers through a cryptographic protocol that ensures illegal modifications can be detected; Coral, a content distribution network to handle the “slash-dot” effect and prevent site collapse; and the LBFS low bandwidth file system, which uses cryptographic techniques to detect content overlap between files to reduce bandwidth consumption by nearly an order of magnitude.
Research Activities
The Stanford Artificial Intelligence Robot (STAIR) project seeks to build a home assistant robot that can do things like tidy up a room. Led by Assistant Professor Andrew Ng, STAIR also revisits the AI dream of building systems that exhibit a significant breadth of intelligence and can carry out many different tasks in a human environment.
Many of us already have robots in our homes, such as our dishwasher, washing machine, and dryer. Andrew believes that a revolution in robotics will come when rather than building such special-purpose robots, we can instead build a single robot that can carry out many different tasks. The STAIR project envisions a robot that can carry out a wide range of tasks such as fetching or delivering items around the office, loading a dishwasher, and preparing meals in a normal home kitchen. Such a robot will revolutionize home and office automation and have important applications ranging from household assistants to elderly and disabled care.
The STAIR robot integrates methods drawn from all areas of AI, including learning, vision, navigation, planning, and speech/natural language processing. This is in distinct contrast to the 30-year-old trend of working on fragmented AI subfields and serves as a unique vehicle for driving forward research toward true, integrated AI.
STAIR is still a young project, but it recently reached its first major milestone of being able to fetch an item in response to a verbal request. A video demonstration can be viewed at cs.stanford.edu/groups/stair/. Faculty members working on STAIR include Andrew Ng (lead), Gary Bradski, Dan Jurafsky, Oussama Khatib, Daphne Koller, Jean-Claude Latombe, Chris Manning, Nils Nilsson, Kenneth Salisbury, and Sebastian Thrun.
In March, Stanford unveiled a new research program: Clean Slate Design for the Internet. The program is a multidisciplinary effort across the departments of Computer Science, Electrical Engineering, and Management Science and Engineering, and the Stanford Graduate School of Business; it takes a nothing-is-sacred approach to improving human communications needs. Unlike other fields of technology, the basic Internet architecture has changed little in the past 40 years. While the lack of change is a testament to the genius of the Internet’s creators, it is also a consequence of the enormous difficulty in evolving and rethinking. The difficulty of bringing about change has held back the networking research community, which has been stuck in “backward compatibility” and “incrementalism” for over a decade. The Clean Slate program’s research agenda can be described by two questions: “With what we know today, if we were to start again with a clean slate, how would we design the network infrastructure?” and “How should the Internet look in 15 years?” You can find more information on this program at cleanslate.stanford.edu.
Seed Funding
This year, the department has started a program to provide seed funding for promising research efforts that are aligned with our strategic plans and are unlikely to receive funding from conventional sources. The initial three projects funded through this program involve research on systems biology, virtual worlds, and robots to assist the elderly.
David Dill and Mark Horowitz have started a project on modeling cellular processes as asynchronous circuits. Their work is based on the idea that living cells must obey many of the same engineering constraints as digital circuits. The processes in cells resemble asynchronous circuits, a somewhat unconventional style of digital circuit that operates without a synchronizing clock. They plan to apply modeling and analysis techniques for asynchronous circuits to complex cellular processes, such as the sequence of steps a cell goes through when it divides. Given an asynchronous circuit model of a process, they can alter a step to model the effect of a mutation and analyze the resulting model to predict the result of that mutation in the laboratory. In the long run, their goal is to develop an engineering knowledge of the fundamental processes of life.
Vladlen Koltun has started a project to build scalable, secure virtual worlds. Virtual worlds are online three-dimensional environments that simulate physical interaction. Virtual worlds have enormous potential benefit, yet little fundamental research has been done. Current virtual world systems are not scalable, not secure, visually poor, and cumbersome. Vladlen and his group plan to remedy this through research on rendering, system architecture, and content. First, they are exploring a hybrid rendering algorithm that decouples per-user bandwidth requirements from the scale and complexity of the world and enables sweeping vistas over vast worlds in real time over a mainstream broadband connection. Second, they are designing a scalable system architecture that enforces fine-grained access controls for privacy and integrity. Finally, they are looking into novel content creation mechanisms that allow casual users to create high-quality virtual world content. The group is building a prototype virtual world system to demonstrate these developments.
Kenneth Salisbury (who also teaches in the Department of Surgery) has collaborated with Laura Carstensen, professor of psychology and director of the Stanford Center on Longevity, to investigate how emerging robotic technologies can be used to improve the quality of life for the world’s increasingly aged population. The effort takes advantage of the complementary interests of the two investigators: Carstensen’s center seeks to bring about profound advances in quality of life for people living longer through scientific and technological breakthroughs that also will benefit people of all ages. Salisbury’s Personal Robotics Project focuses on the creation of robots that are “human compatible” in the mechanical sense (range of motion, energy and impact bounds, impedance, texture, operational in wheelchair-accessible environments, etc.) as well as in the behavioral sense (personality, initiative, judgment, etc.).
Educational Affairs
To help our undergraduate program evolve as quickly as the field and to maintain our leadership role in CS education, we have embarked on a significant restructuring of our undergraduate curriculum. The Curriculum Committee, including Alex Aiken (chair), Jerry Cain, Bill Dally, Daphne Koller, Vladlen Koltun, John Mitchell, Nick Parlante, Bob Plummer, Eric Roberts, Mendel Rosenblum, Mehran Sahami, Claire Stager, and Julie Zelenski, has helped create a new structural outline for our undergraduate major that provides students with more options to specialize their programs, take more depth courses in areas of interest, and engage in meaningful multidisciplinary work. Initial response to the proposed changes by the external visiting committee has been positive. In the coming year, we plan to continue with the proposed curriculum redesign.
As part of the transition of CS106A to Java, Eric Roberts’ textbook The Art and Science of Java was released earlier this year, creating a national imprint for our ongoing work in undergraduate curriculum development. Also, under the organizational leadership of Jerry Cain with IT support from Miles Davis, we helped host the 2006 ACM Pacific Northwest Programming Contest at Stanford. In a very competitive region of 65 teams, Stanford did quite well, with our three teams placing third, fourth, and ninth, respectively. The third place team, comprising James Connor, Justin Solomon, and Alex Utter, and coached by Jerry, advanced to the world finals in Tokyo, Japan, where they tied for 26th place in a field of 88 qualifying international teams. We are proud of their performance, which included besting Cal, our friendly rivals across the bay, in the regional contest.
Computer Forum
The Computer Forum is approaching its 40th year as the Industrial Affiliates Program for the Department of Computer Science and the Computer Systems Laboratory. The Forum continues to offer many opportunities for our faculty, students, and industrial partners to work together to advance research and focus on real-life problems such as computer security, with an ultimate goal to place our students into highly regarded high-tech companies.
Under the leadership of Professor John Mitchell and Suzanne Bigas, the 60-plus corporate members of the Forum provide a reliable source of income to support our students and research activities in the CS department and CSL. The Computer Forum helps Stanford maintain its world-class reputation in computer science and sponsors a variety of activities, including internships, research funding, jobs, opportunities to collaborate, and student support.
Awards and Honors
Faculty
Mark Horowitz and Sebastian Thrun have been elected to the National Academy of Engineering. This is one of the engineering profession’s highest honors. Mark was lauded “for leadership in high-bandwidth memory-interface technology and in scalable cache-coherent multiprocessor architectures.” Sebastian was honored “for contributions to probabilistic robotics, including mobile robot localization and mapping.”
Bill Dally and Pat Hanrahan have been elected Fellows of the American Academy of Arts and Sciences. An independent policy research center, the academy undertakes studies of complex and emerging problems. Current academy research focuses on science and global security, social policy, the humanities and culture, and education.
Dawson Engler has been awarded the 2006 ACM SIGOPS Mark Weiser Award, given to an individual whose contributions in operating systems research are highly creative, innovative, and possibly high-risk.
Pat Hanrahan has been awarded the 2006 IEEE Visualization Career Award. This award recognizes Pat’s lifetime of contributions to visualization research, spanning such diverse areas as volume rendering, scientific visualization, and information visualization.
Vladlen Koltun and Andrew Ng have received Sloan Research Fellowships. Established in 1955 by the Alfred P. Sloan Foundation, the fellowships recognize and support young faculty who show outstanding promise of making contributions in specified fields of science. This year, 118 fellowships were awarded in the fields of chemistry, computational and evolutionary molecular biology, computer science, economics, mathematics, neuroscience, and physics.
Subhasish Mitra has been awarded the NSF Faculty Early Career Development (CAREER) program award. It is the National Science Foundation’s most prestigious award in support of the early career development of those teacher-scholars who most effectively integrate research and education within the context of the mission of their organization.
Staff
Claire Stager and Isis Contreras have received Computer Science Distinguished Service Awards. This is an annual award presented to two staff members for their extraordinary service to the department.
Students
Yi-Ren Ng was awarded the ACM Doctoral Dissertation award for his dissertation, Digital Light Field Photography done under the supervision of Professor Pat Hanrahan. Yi-Ren has founded Refocus Imaging to commercialize his research on a new kind of digital camera that takes photographs that you can refocus after the image has been taken.
Stanford beat Cal in the Google Puzzle Match to solve puzzles and build bridges of Lego blocks.
Staying Connected
Please visit our website regularly at cs.stanford.edu for updated information about the department. From this site, you can link to many sources for detailed information about our faculty, students, research programs, and teaching initiatives. I welcome your suggestions regarding the department’s directions and activities, and encourage you to visit us to see firsthand our facilities and programs.
I hope that you will remain an active member of the department’s alumni community by keeping us apprised of your activities and whereabouts. You can log on to soe.stanford.edu/alumni/ update.html to update your contact information. Or, if it is more convenient, fill out the enclosed alumni news update form. Thank you for keeping in touch.
With best regards,
William J. Dally
William R. and Inez Kerr Bell Professor
Chair, Department of Computer Science
- Login to post comments