Intelligent Machines

Antarctic Ice Divers

Two robots have gathered a treasure-trove of data from beneath the ice shelf.

May 13, 2011
Last November, a team of scientists, funded by the National Science Foundation, began using a pair of autonomous robotic vehicles to gather an information from beneath the Ross Ice Shelf in the Antarctic.
The team used two iRobot’s 1KA Seagliders for the two-month mission. The robots can dive down to 1,000 meters underwater for eight to 10 months at a time. They do this using a “buoyancy engine,” which consists of a pump that moves oil to and from an external bladder into a reservoir to change the robot’s density.
Monitoring ocean temperature, salinity, and phytoplankton activity can helps scientists better understand glacial ice melt, which can reveal more about climate change. But gathering data in such icy, harsh conditions is extremely difficult; scientists usually need to rent out ships and can only take a handful of samplings from the water over a few days.
Two team members under Vernon Asper, a professor of marine science at the University of Southern Mississippi’s Stennis Space Center, and Walker Smith, professor of marine science at Virginia Institute of Marine Science, prep one of the gliders for launch.
The robots navigated using compass readings and by keeping track of their speed and movement relative to their last GPS reading. To change their direction, they shift an internal bank of batteries.
When a Seaglider surfaces, it uses its antennae to send data and position information back to the team via satellite, and to receive commands. Seagliders have been used to monitor oceans and marine life elsewhere, and were used to track the oil plume in the Gulf last year.
This image was captured from beneath the ice not by one of the robots but by a member of the scientific team.
This image shows the path the two gliders took across the Ross Sea. A heavy ice patch is visible to the left, and cloud coverage to the right. The team is still processing and analyzing the data collected by the Seagliders, and will use it to model different ocean-related processes over time. “This is the first time we’ve been able to put autonomous equipment in the water and basically go home and let it send us the data,” says Asper.