By Rob Knies
December 8, 2008 11:58 AM PT
Microsoft Research Asia will be celebrating its 10th anniversary during the first week of November, and to mark the occasion, the lab is hosting a collection of events designed to engage renowned technologists, the academic community, and its own researchers. In preparation for the festivities, Hsiao-Wuen Hon, managing director of the lab, took a few minutes to look back on Microsoft Research Asia’s first decade—and to discuss the one ahead.
Q: It’s been a year since you took over as managing director of Microsoft Research Asia. What have you achieved over that time?
Hon: I’m very pleased with the results in both academic research and tech transfer—and also with our collaboration with Asia Pacific academics. We had probably the best year in our 10-year history in terms of tangible results.
In many top conferences, such as SIGGRAPH, SIGIR, WWW2008, and ACM Multimedia, researchers from this lab have contributed close to 10 percent of the accepted papers. We also won several best-paper awards in prestigious conferences. In the systems research area, we have started to get into the top conferences, such as OSDI. I’m very pleased about that.
In search and advertisements and Office and Xbox, we’ve had significant technology transfers. It’s a tribute to the entire lab for making some long-term bets. Many of our projects have started bearing fruit after years of hard work, so it’s thanks to the researchers and the leadership in Microsoft Research Asia in the last few years that we have achieved such success.
At the same time, this year is our exciting 10th anniversary, and we are really looking forward to the week. We will have more than 100 internal Microsoft guests, including many of our colleagues in Microsoft Research and partners in product groups. We will make sure it’s a productive and fun event.
During the Summer Olympics, when Bill Gates visited, we held a very successful Innovation Forum in Hong Kong. It was well covered by the media, and we were particularly excited to have an opportunity to share our experience with the Asia Pacific region and the world, to show what this lab has done in the past 10 years and to promote the vision we have. It was a good opportunity to reflect on the past and, at the same time, look forward to the future.
Q: What are the coolest things your lab is working on at this moment?
Hon: Our work is categorized into five broad areas: First is to develop the best user interface, to have a more natural, better user experience. We invest a lot in speech, handwriting, natural-language, and computer-vision technology, to build a computer that can listen, can see, and can understand. We have been making good progress in all these areas that has or is in the process of getting into the next versions of Windows and other Microsoft products.
The second category is multimedia technologies. This is the area we are most known for in the first 10 years, and we have produced many exciting multimedia technologies that have gone into Windows, Xbox, the Entertainment and Devices Division, and the Live Search product and services. One of our top researchers in this area is Xian-Sheng Hua, who was just named to Technology Review’s TR35. Many of our video technologies are very promising for the next generation of multimedia applications.
In data-centric computing, with the emerging focus on cloud computing and software as a service, there’s plenty of opportunity and need for highly scalable infrastructures and analysis tools to process today’s massive amount of data and to turn that data into information and knowledge. The many Internet service providers can use that data to provide a targeted service. For the enterprise, business intelligence is an emergent area, and in Microsoft Research Asia, we are excited to engage partners in trying to build the live, data-centric computing infrastructure and tools and applications to get involved.
Search and ads represent a powerful combination of technical and business models for delivering Internet services. We have invested a lot in the last few years into work with the Live Search and adCenter teams to provide the best services and technologies to compete in this space.
Also, in computer-science fundamentals—theory, systems, networking, and other emerging research methodologies (what we call Research 2.0) that are serving as the foundation for computer-science research—we have invested significantly in systems and networking. As I mentioned, we are starting to participate in OSDI with high-quality papers, and that’s refreshing.
Q: When you review the lab’s history over its first 10 years, what are the highlights?
Hon: Three things. The first is innovation. We are very proud we have extended the vision Microsoft Research Senior Vice President Rick Rashid and Bill Gates established 17 years ago into China and Asia. We’ve built a world-class research lab, fostered an open, free environment, hired the best people, and innovated. In area such as computer graphics, computer vision, multimedia, and Internet search, we have demonstrated that we can do world-class research. Some of the achievements are beyond the expectations of anyone, myself included, who witnessed the opening of this lab 10 years ago.
Second is talent. The fundamental reason we choose a place to start a research lab is talent. There’s no doubt China produces a lot of engineers, but 10 years ago, no one knew what their quality was, particularly when we talk about people who can do world-class research within a short time, the best of the best. I think we proved we could find that top talent and give them an environment in which to succeed.
What we are most proud of are our own people. Many of the people we hired 10 years ago, locally and just out of university, have become our senior researchers and research leaders. They can lead many junior researchers we’ve hired in the last couple of years to produce world-class research. In talent development, we certainly feel we have achieved our most important goal and will be able to continue our success for many years to come.
We are very proud of our internship program, which we consider unique. We have hosted more than 3,000 interns in the last 10 years. The numbers are just mind-boggling. Many of our interns went on to top universities such as MIT, Stanford, and Carnegie Mellon. Microsoft Research Asia has become a recognized brand for interns applying for Ph.D. programs in top universities worldwide or for professional jobs. That’s recognition of the quality of our internship program.
The third highlight is collaboration. Like other Microsoft Research labs, we have a University Relations external research organization that works with lab researchers on collaboration with universities. Most of the papers we produce are collaborations with university faculty and students. We have a very broad and deep collaboration with Asia Pacific academia to promote research, curriculum, and talent development. We also run a lot of academic exchanges so that people can come together to share their experiences and best practices.
We set the model for how to run university collaboration in the Asia Pacific region, particularly in China. I think all the multinational companies in China copy our model of doing university relations and collaboration, and we are very proud of that.
Q: What’s it been like for you personally to move back to Asia and into a role where you’re heading a world-class research facility?
Hon: I was born in Asia, grew up in Asia, and then came to the U.S. to get my Ph.D. I worked in the U.S. for almost 15 years and then returned to Asia. My personal experience is kind of analogous to what most Asian countries went through in the last 20 or 30 years.
Over that span, there are roughly three periods. The first is the learning period. We’re trying to learn science, technology—all the good things from the West. For me, I went to Carnegie Mellon to get my Ph.D.: a learning period.
Then we went into the next phase, the experimenting period. I started working in the U.S., and, just like Asia Pacific countries—whether in science, technology, or economic reforms—tried to improve and grow.
The third period is to mature and start bearing fruits for our own interest. For instance, China is by far the No. 1 mobile-phone market, and just this year, China bypassed the U.S. as the No. 1 country in Internet usage. Korea and Japan also have enjoyed their own successes.
At Microsoft Research Asia, we have been fortunate to share in that third period of the transformation. We have grown along with the region, and this is certainly one of the key factors for us to be able to achieve what we have in the last 10 years.
I am excited to come back to Asia and work with people in the Asia Pacific region. To be able to come back here and contribute to the Asia Pacific society and region has been a very rewarding, stimulating experience. The world is flat. I have so many friends—American friends, Western friends—come to China and Asia. At the same time, there are many Asian people continuing to go to the States. We all collaborate on a worldwide basis.
Q: Looking ahead to the next 10 years, what are the areas on which you expect your lab to focus?
Hon: Software plus services is Microsoft’s technology road map, and I think the five categories I mentioned earlier are nicely aligned in this direction. The client will continue to be extremely important for Microsoft’s strategy, because to have a great software-plus-services offering, you need to have a powerful client, whether it is in a PC, a mobile device, or other personal devices.
With the best user interface and with multimedia, you’re trying to provide the user with a great user experience. It’s important for Microsoft to continue to provide the best client experience. There is legitimate competition in both the PC front and the mobile-phone front, and there’s a tremendous opportunity for us, in user interfaces and in multimedia technology, to make sure we continue to be a leader in the client area.
With regard to data-centric computing, search, and ads, and that’s trying to provide the infrastructure and tools, and to build the best service in the cloud. Cloud computing is an interesting area, one in which the company has a big investment. We actively engage with many partners in this technology, and I think the opportunity is tremendous.
We will continue to pursue the fundamental technologies: theory, systems, and networking—and a few Research 2.0 methodologies. There are two things we are trying to promote across the whole lab: treating the Web as a platform and using social computing to help our research.
Another is deployment-driven research. We have a new initiative, called MSR.CN, in which we are trying to push out beta versions of our research, get early feedback, and close the feedback loop to make sure we can improve and speed our innovation cycle. We’ve started those kinds of initiatives in the last couple of years, and researchers have embraced this idea. In some of these areas, we already see early benefits.
Q: Ten years from now, how do you want people to describe Microsoft Research Asia?
Hon: First and foremost, I want them to continue to think of Microsoft as an innovator. We have very fierce competition from high-tech companies and people generating new technologies. We cannot sit still.
Research is at the forefront of making sure Microsoft has a future.
The second thing would be that people recognize Microsoft Research’s contribution to Microsoft. I think the best way is through the technology we develop that goes into Microsoft products and the mind share we have in the industry and the academic front.
In terms of this lab, Microsoft Research Asia, we are part of Microsoft Research family. Yes, we have our own unique characteristics, as we have demonstrated in the first 10 years, and in the second 10 years, we definitely would like to continue many of our great legacies of the first 10 years.
At the same time, there are a lot of things we need to learn from our sister labs, from Microsoft, from the whole world. The Asia Pacific region is exciting, but we live in a much smaller world now, a world that is all connected. We want to provide a unique contribution—there are plenty of collaborations and plenty of things we can learn—and we also want to maintain a humble, confident mentality that we can continue to innovate, continue to do more and better.