Microsoft in India
A top executive at Microsoft Research’s newest lab, in Bangalore, India, outlines its evolving R&D strategy.
As Microsoft’s products have spread around the world, so has the company’s 15-year-old R&D wing, Microsoft Research (MSR). In addition to its original location at Microsoft’s main campus in Redmond, WA, the 700-member research division has established labs in San Francisco; Silicon Valley; Cambridge, England; Beijing, China; and most recently, Bangalore, India.
The Bangalore lab, which opened in January 2005, represents the company’s gambit to tap into India’s rich university research community and considerable software engineering talent. Bangalore, capital of the state of Karnataka in southern India, is home to hundreds of information-technology companies, including two of the country’s three largest software firms, and is often called “India’s Silicon Valley.” Thus, for U.S. technology companies to be putting software labs in Bangalore is, in many respects, a no-brainer. Indeed, IBM, Hewlett-Packard, Intel, and other corporations already have labs there.
[For images of children in MSR’s program, click here.]
But a presence in India may be especially important for Microsoft. Not only does the company hope to put its Windows and Office software on a larger share of computers in India – where desktop PC sales grew more than 20 percent between March 2005 and March 2006 – but it sees the country’s billion-plus population as a market for future Microsoft products, such as educational software specialized for schools in poor, rural communities.
Last week, MSR India’s assistant managing director, Kentaro Toyama, was in Berkeley, CA, coordinating an international conference on information and communications technology for developing economies. Technology Review Senior Editor Wade Roush caught up with him there.
Technology Review: Why is it important to Microsoft to have a software lab in India?
Kentaro Toyama: India provides a unique environment for certain kinds of research. It has a booming IT economy while at the same time having a large percentage of the population still in poverty. So there are certain kinds of research that would be difficult to do in the United States that we can do there, particularly with respect to the role of computing in poor communities. And there is a well-educated base of computer scientists and engineers, including a returning diaspora, if you will, of Indians who have worked overseas.
TR: What kinds of research are you starting with?
KT: The six research areas we’re currently focusing on are cryptography*; digital geographies, which includes any kind of digital map or location-based services and software; multilingual systems, including speech recognition, natural language processing, and building systems that interact across different languages; hardware for communications, including distributed sensor networks; software engineering, which looks at creating tools that make software development easier; and emerging markets, or how computing will impact social and economic development.
TR: Say a bit more about multilingual systems – India would seem to be the perfect place for that kind of research.
KT: Yes, India is an amazingly diverse country in terms of languages. There are 22 officially recognized languages and hundreds of other languages and dialects. The average person on the street is bilingual and often trilingual. They’re quite comfortable switching back and forth. But most computer interfaces are for people who only speak one language. We’re very interested in supporting technologies for multilingual use of the computers. For example, if you’re doing a search online, you want to be able to return results that are well adapted to the users’ languages.
TR: What specific problems are you looking at in the area of software engineering techniques?
KT: In India a lot of the software development that happens needs to be coordinated with efforts in the U.S. or other countries. Usually these places are thousands of miles away and there are problems of coordination. The question is: What are the actual problems – the real issues when you’re trying to do programming across continents? Is it possible, for example, to write specifications for a particular program such that you can really just throw it over the fence [to a team in a different country] and it doesn’t require a lot of discussion?
TR: You’re in Berkeley leading a conference on community technology for developing countries. How does MSR India’s own work fit into that theme?
KT: We are interested in finding ways for computing to make an impact in rural villages or in urban slums. You can go and do user studies with people who live in these communities, and because India has a fairly good IT infrastructure, all of the hardware and software is right there across the street. And there is a lot of interest in India in doing work like this. I’m surprised by the sheer number of nonprofits and NGOs [nongovernmental organizations] working in this area. There is a cultural inclination toward doing this kind of work.
* Correction, June 2, 2006: Due to a transcription error, the original version of this story listed photography as one of the Bangalore lab’s main themes, rather than cryptography.
TR: What types of community projects are you working on?
KT: We initially wanted to see if there was something we could do for rural schools that would help the educational system. We visited a number of schools, quite a few of which already have PCs. In fact, it turns out that 5 to 10 percent of the rural schools in India have PCs. But they are not able to afford a ratio of one PC per child. So every time we would see the same situation: one PC surrounded by five or 10 kids, and usually a dominant child in the center, dominating the mouse and the keyboard, while the other kids are not benefiting as much. This is a situation where we can multiply the value of the PC by adding other kinds of interaction. We developed software for PCs that will support as many mouses as you have USB ports. Each mouse has a different-colored cursor, allowing, say, competitive flash-card games that encourage certain kinds of learning. We’re doing preliminary studies to see if there is educational value. We believe we can take a situation where there is one kid and he’s having a great time – and it’s almost always a “he”, and the other kids are fighting over the mouse, or are simply bored – and turn that into a situation where all of the children are thoroughly engaged.
TR: It’s hard to imagine ten school kids bunched around a single computer, all interacting with it at the same time. Does it really work in practice?
KT: Somewhere between six and ten kids, this model breaks down, but up to five is quite doable. The software does need to be specially written for this paradigm. We’re hoping to release a software development kit so that people can write educational content using multiple mouses. We’re expecting to pilot the technology in schools in and around Bangalore, and if those work out, we’d like to scale up the study, not just in India but beyond.
TR: How does having a lab in India help Microsoft in the short term?
KT: Corporations are really waking up to the potential business in emerging markets, and as a result there is growing interest across Microsoft in understanding countries like India and China much better. We gather a lot of knowledge just being on site and understanding everyday life in India. When we come back to Redmond we spend a lot of time with product groups, some of which are very explicitly interested in emerging markets, and some of which are just considering the idea. In each case we are still in the early learning stages. We are trying to understand the people who are fairly wealthy and are probably our first, immediate customers, and then trying to see how we can make an impact [among other demographics] where we may not see any short-term revenue, but where, if you engage, you will open up new markets in the future.
TR: Since September 11, it’s often been difficult for foreign scientists and engineers to get the necessary visas to work in the United States. Does having a lab in India help Microsoft get access to researchers you might not be able to employ otherwise?
KT: In theory, that would be a great idea, as a way to bring in people who are very good technically but who cannot come to the U.S. for whatever reason. But actually, most of the people we’ve hired have very voluntarily either returned to India from abroad or stayed in India. We are tapping into a population that really has the option to work more or less anywhere, but they are by choice staying in India.
Couldn't make it to EmTech Next to meet experts in AI, Robotics and the Economy?Go behind the scenes and check out our video