Microsoft Research collaborates with academic, government, and industry researchers across the globe on projects that support environmental initiatives—such as helping Indian farmers protect their land and helping scientists develop new tools to analyze the large volume of environmental data.

Where Great Scientists Follow their Instincts

In 1991, Microsoft became one of the first software companies to create its own computer science research organization. Microsoft Research has developed into a unique entity among corporate research labs by balancing an open academic model with an effective process to transfer its research to product development teams.

With 800 researchers and laboratories located in Washington state, California, Massachusetts, Bangalore, Beijing, and England, Microsoft Research is recognized worldwide as a premier computer-science research institution.

Kevin Schofield, general manager for Microsoft Research, believes Microsoft Research's ability to find the best possible partners results from its willingness to let great scientists follow their instincts: "It's a bottom-up process," he says. "You can think of Microsoft Research as the world's largest computer science department."

"We hire really smart researchers, they hire junior researchers to build out their team, then they decide what technologies they want to pursue and who in the academic world they'd like to partner with. And we don't tell them what to work on."

Microsoft Research collaborates with scientists around the world on a wide range of environmental research projects. Support includes funding, software, and technical and research expertise. The result of that collaborative approach is that the research that emerges from Microsoft Research reflects local priorities and meets local needs.

Digital Green: A Local Approach with the Right Technology

Microsoft Research's work with local partners allows its scientists to learn how technology can help with local needs. Consider the Digital Green project, launched in 2006 by Microsoft Research India's Technology for Emerging Markets team.

Fast-growing India faces the challenge of food production for more than 1 billion people. Many of the nation's farmers struggle with limited access to effective and sustainable farming techniques.

The Digital Green (DG) project is developing a participatory framework for agricultural extension through digital video. The DG project has delivered 150 videos that demonstrate helpful farming techniques to remote farms and villages near the Karanataka and Tamil Nadu-state borders. Many of these videos feature farmers who are known to those who watch the videos The videos demonstrate how farmers can boost crop production, farm sustainably, and build new markets for their products.

The Digital Green videos are local and relevant. The DG system provides structure to the traditional, informally-trained vocation of farming. The system improves the efficiency of extension programs by delivering targeted content to a wider audience and enabling farmers to better manage their farming operations with reduced field support.

Familiar Faces, Relevant Help

The DG project adopted easy-to-use digital video cameras that filmed local farmers and agricultural experts practice and discuss improved farming techniques such as adopting mixed crops, reducing pesticide use, and producing compost.

The videos were minimally edited, then distributed and shown to individuals or small groups on notebook computers, portable DVD players, or in some cases through a village's cable TV network.

"The people who watch the videos find them more credible because they see a farmer they know who has the same problems they do and has adopted these practices," says Schofield.

In time, DG may become an independent non-governmental organization (NGO) and expand its scope across India.

Digital Watershed: Collaborating to Analyze an Ocean of Data

"Scientists are drowning in data today," says Schofield. "Through our partnerships we hope to understand the advances in computing that we need to make to keep up with the needs of scientists."

One such collaboration in the United States includes a number of "digital watersheds" built by computer scientists and scientists at Microsoft Research's eScience Group, Berkeley Water Center, and Lawrence Berkeley National Laboratory, and National Marine Fisheries Service.

The collaboration's goal is to make efficient use of advanced but sometimes incompatible sensor networks to create a detailed picture of what occurs in a watershed such as that of the Russian River, which runs through the wine country of Sonoma County, California and is threatened with over-use, pollution, gravel mines, and more.

The Russian River digital watershed explores management policies. Lawrence Berkeley researcher Deb Agarwal worked with Microsoft Research member Catharine van Ingen to create a database that could organize data that was extensive but almost incomprehensible because of its wide variety of sources and measurement methods.

Next, researchers will attempt to build a data map of the Russian River by layering data from real-time sensors with historical California river records (some of which are more than 100 years old) to capture perhaps the most detailed image of a river ever created. The project is expanding to all over California and includes almost 4000 historic sensors and new sensor deployments.

The Digital Watershed project has more than the potential to help improve water management. It's an example of work by Microsoft Research and partners to organize and analyze data that is the result of high-quality, low-cost data collection and storage, and increased use of sensors and computer hardware as data-collection tools.

Other Partnerships Address Key Global Issues

Elsewhere, Microsoft Research partners with the Switzerland-based Swiss Experiment, a collaboration of environmental science and technology research projects that aim to answer fundamental questions about changes in the natural environment.

"Microsoft Research is supporting the Swiss Experiment project to help find a solution for the problem of global warming through the application of emerging sensor network technologies and the creation of a platform where valuable sensor information can be shared," said Tony Hey, corporate vice president for External Research, Microsoft Research. "Our goal is to provide tools to help researchers address some of the toughest environmental issues facing the world today."

A variety of environmental projects at various sites across Switzerland will be able to use this data infrastructure to improve models that predict natural hazards such as floods, avalanches, and landslides, as well as to document environmental degradation and change.

In Russia, Microsoft Research is supporting research projects that provide a unique opportunity to apply high-performance computing and Microsoft technology to study issues that range from the impact of regional climate change on various ecosystems on the vast territory of Northern Eurasia to Saint Petersburg's flood defense. Novel computational methods and tools developed through these projects are expected to benefit not only the Russian territory, but contribute to the broader international effort.

The Climate Induced Vegetation Change Analysis Tool is a collaborative research project with Space Research Institute (IKI) and Geophysical Center (GC) of the Russian Academy of Sciences (RAS) and Microsoft Research focused on merging large archives of satellite images with historical data on vegetation and climate for the territory of Northern Eurasia.

Another project in that region is Microsoft High Performance Computing Technologies for Safety and Environment (MicroTEST)—a two-part project with the Laboratory of Applied Mathematics and Mechanics at Saint-Petersburg State Polytechnic University. The first project looks at the development of a new efficient computational methodology that can predict the interaction of fine water spray with buoyant turbulent diffusion flame occurring in fires. The second project focuses on detailed mathematical simulation of the dynamic behavior of the gates of Saint Petersburg flood-defense system ship-passing channel under pressure from moving water.

Other projects include a large-scale environmental study project that deploys hundreds of miniature weather stations in schools throughout Singapore. And in the recent Power Aware Computing Request for Proposals, Microsoft Research focused on power consumption in data centers in the United States. To encourage novel ways to think about how to reduce data center power consumption, Microsoft Research offers USD $500,000 in grants to research energy efficiency innovation in data centers.

Advancing Computer Science to Solve Difficult Problems

Microsoft Research and its partners are breaking new ground in using the power of technology to solve fundamental environmental problems. Schofield sees universal benefits.

"We're getting a huge benefit at Microsoft in advancing the state of computer science by taking on these hard science problems," he says. "And scientists get real benefit from having experts in computer science advance the state of the art. It's a great match."