Grant Will Fund Wireless Environmental Sensor Network

November 8, 2005

Network enables researchers to install sensors in remote locations and collect data automatically for years

Duke University environmental scientists have received a five-year, $1.88 million grant from the National Science Foundation (NSF) to develop an advanced wireless sensor network that can measure, model and predict biophysical changes in the forest environment.

The network will help researchers better understand how the growth, survival and reproduction of forest trees are influenced by changes in climate, atmospheric carbon dioxide and other environmental variables that can fluctuate rapidly. Such changes are expected with the ongoing alteration in global climate as increasing carbon dioxide levels from burning fossil fuels cause global warming.

James S. Clark, who is H.L. Blomquist Professor of Biology at the NicholasSchool of the Environment and Earth Sciences, is principal investigator on the grant.

“This network will allow us to go into remote locations, install the sensors, and, for years to come, collect a depth and breadth of data that would be virtually impossible to obtain through any other means,” he said. “It has the potential to let us study environmental change on a whole new scale.”

Clark's co-investigators are Alan Gelfand, James B. Duke Professor of Statistics and Decision Sciences; Pankaj Agarwal, Earl McLean Jr. Professor of Computer Science and Mathematics; Carla Ellis, professor of computer science; Kameshwar Munagala, assistant professor of computer science; and Jun Yang, assistant professor of computer science.

They will collaborate with Paul Flikkema, professor of electrical engineering at NorthernArizonaUniversity. Flikkema received an additional five-year NSF grant for $760,000 to work with the Duke team on the network.

Judson Edeburn, resource manager for DukeForest, is also assisting the team.

A key function of the proposed network will be its ability to automatically regulate its own performance and, when necessary, adjust how and when data are collected. By combining their scientific and engineering expertise, the Duke-Northern Arizona team plan to build in ‘smart’ technology that will allow the network to assimilate large amounts of data in real time, rapidly assess the value and cost of collecting the data, and then automatically schedule future data measurement accordingly, Clark explained.

This self-managing capability will make the network a more efficient tool for collecting environmental data in remote locations where there is little or no access to traditional power sources and other infrastructure, he said. It also will enable sampling data over intervals that might range in time from seconds to years, and in space from meters to entire landscapes measuring miles.

Once collected, the data will be relayed back to the network’s server, where it will be assimilated, analyzed and stored for use by scientists, conservationists, land managers and teachers.

The network’s control model will be updated periodically to reflect new knowledge of environmental variables and network parameters, such as battery life.

The new network builds on successful past efforts by Clark and his colleagues to construct and deploy wireless environmental sensing networks in two forests in Piedmont and western North Carolina. The palm-sized, weather-resistant sensors they developed can be placed virtually anywhere in a forest -- high in the tree canopy, in the understory or near the forest floor. Hundreds of them can be deployed to “network” an entire forest.

As the new technology develops, Clark believes it will become a powerful new tool, not only for research, but also for conservation, forest management and environmental education at the graduate, undergraduate and high school levels.

URL for complete story: http://www.dukenews.duke.edu/2005/11/wirelessnetwork.html