Click here to bypass page layout and jump directly to story.=


UC Berkeley >


University of California

Press releases Top stories News - Media Relations

Berkeley








NEWS SEARCH



NEWS HOME


ARCHIVES


EXTRAS


MEDIA
RELATIONS

  Press Releases

  Image Downloads

  Contacts


  

 Press releases

 
Wireless sensor mote
More than 20 miniaturized wireless sensor motes, such as the one pictured above, are now sending back raw data on the microclimate of a petrel seabird colony in Maine. This newest generation of motes, called Mica, could help change the way conservation biologists monitor wildlife habitat. (Photo courtesy Intel Research Laboratory at Berkeley)

Wireless sensors from UC Berkeley and Intel researchers help conservation biologists monitor elusive seabird in Maine
05 August 2002

By Sarah Yang, Media Relations

Berkeley - For scientists studying the Leach's Storm Petrel, monitoring the shy seabird's nest activity meant sticking a cumbersome remote camera or a daring arm into burrows. But starting Monday (Aug. 5), biologists and petrel buffs around the world will be able to monitor a popular breeding site in real time through the Internet while sitting comfortably in front of their computers.

Wireless sensor technology developed through a partnership between the University of California, Berkeley, and the Intel Research Berkeley laboratory is making this powerful method of habitat monitoring possible. Over the summer, researchers from the Intel lab, in the city of Berkeley, and from UC Berkeley teamed with biologists from the College of the Atlantic in Maine to install a network of more than 20 miniaturized sensors, or motes, on nearby Great Duck Island.

Each device, slightly bigger than the two AA batteries powering it, is now beaming back raw data about the conditions in the burrows and the island's microclimate that will be viewable to the public at http://www.greatduckisland.net.

"There is nothing else like this sensor network available for conservation biologists, nothing that can provide good quality data in such dense numbers," said John Anderson, associate dean of advanced studies and conservation biologist at the College of the Atlantic. Anderson and students from the small liberal arts college in Maine have been studying seabird colonies on Great Duck Island for the past four years. "What's really exciting about this is that we can get a feel for what happens on the island when humans aren't there," he said. "This kind of sensor network will have a profound effect on how we do field ecology."

The raw information provided by the motes will help biologists understand why the Leach's Storm Petrel (Oceanodroma leucorhoa) favors Great Duck Island over thousands of other islands off the coast of Maine. "This is particularly important in conservation questions," said Anderson, who also is a UC Berkeley alumnus. "We want to know which islands are important for the seabirds, and why. We could learn that the microhabitat of Great Duck Island is unique and in need of protection."

  Leach's Storm Petrel
A graduate student examines a Leach's Storm Petrel on Maine's Great Duck Island. (Photo courtesy of John Anderson)
 

Up to now, obtaining an accurate count of the elusive seabirds involved expensive, carefully planned trips to the island with a pen, paper and a portable video system - dubbed the "petrel peeper" - that was transported by wheelbarrow or several biology students.

Biologists believe the 237-acre island, located 12 miles from Acadia National Park, may be home to one of the largest petrel breeding colonies in the eastern United States. The elusive seabird is particularly challenging to study because it spends most of its life offshore, returning to land only during the sensitive breeding period from the end of May through October. On land, the petrels stay hidden during the day to avoid predators, typically emerging after 10 at night.

The motes were placed in six burrows and the surrounding brush, covering an area just larger than half a football field. "From the biological side of things, sample size is very important," said Anderson. "The more individuals you can look at, the more you know about the health of the colony."

The idea for the project began through Anderson's friendship with Alan Mainwaring, a research scientist at the Intel Research Berkeley laboratory. "We were looking for an application to field test the motes, and John needed a reliable way to monitor the petrels throughout the year," said Mainwaring, who received his PhD in computer science from UC Berkeley three years ago.

With Anderson's input, the researchers designed the motes to detect light, barometric pressure, relative humidity and temperature conditions. An infrared heat sensor detects whether the nest is occupied by a seabird, and whether the bird has company. Motes within the burrows and around the brush send readings out to a single gateway sensor above ground, which then relays collected information to a laptop computer locked away at a lighthouse on the island. The laptop, also powered by photovoltaic cells, connects to the Internet via satellite.

The sensors used in this project are the newest generation of UC Berkeley motes, called Mica, and communicate with each other via radio signals sent at 40 kilobytes per second. David Culler, computer science professor at UC Berkeley and director of the Intel Research Berkeley laboratory, worked with UC Berkeley graduate students to develop the sensor boards and networking software for the Mica motes.

"The unique requirements of monitoring wildlife really pushed the engineering of the Mica motes in new directions," said Culler. Biologists aren't able to visit research sites more than a few times a year, and they avoid doing so to lessen the negative ecological impact of repeated human presence. It therefore became especially important to reduce the motes' power consumption so the sensors would last between visits. The motes in the burrows can run continuously on two AA batteries for six months, while the relay sensor - which consumes more power - is rigged to a solar panel.

Biologists also have the flexibility of moving the sensors around to where they are needed most. "The network is self-organizing," said Culler. "The nodes will automatically search for signals from neighboring networks and adapt to changes in position."

Joe Polastre and Robert Szewczyk, two UC Berkeley graduate students and summer interns at Intel, noted the importance of keeping the motes small enough so they wouldn't be intrusive for the seabirds. "We can even shove the motes into the dirt walls of the burrow if we need to," said Polastre.

Mainwaring pointed out that motes placed in an office building or home do not need to contend with rain, sand or a curious seabird's beak. He said the motes in the burrows are protected from dirt and moisture by a 10-micron thick polymer coating that is biologically inert. An acrylic, cylindrical casing protects the motes above ground.

The researchers will be monitoring the data and improving on the system over the next several months. They plan to return to Great Duck Island in April to deploy more motes before the start of next year's breeding season.

The collaborative project is one of many innovative initiatives associated with the UC Berkeley-based Center for Information Technology Research in the Interest of Society, or CITRIS. The center is one of four California Institutes for Science and Innovation launched by Gov. Gray Davis to spark a new generation of cutting-edge technologies.

Through CITRIS, the researchers are developing future habitat monitoring projects at sites within the UC Natural Reserve System, including the James San Jacinto Mountains Reserve in Idyllwild, Calif. The researchers also have many ongoing projects with other UC Berkeley departments using sensor networks.

The project also gives computer scientists a chance to get away from their desks and into the field. "The thing that really attracted me to this project was the ability to take abstract algorithms and try them out in a real application," said Robert Szewczyk. "We've got these small devices that we're claiming can be deployed for remote sensing applications. So let's try it, right?"

###