3 August 2006

Truly automated robots: No humans allowed

A new mathematical system that allows undersea robots to self-choreograph their movements in response to their environment may one day power, with no human supervision, other robotic teams that could explore not just oceans, but deserts, rain forests, and even other planets.

One move toward that end: This month an entire fleet of undersea robots will for the first time work together without the aid of humans to make detailed and efficient observations of the ocean in Monterey Bay, Calif.

The goal of this mission is to find out about a periodic upwelling of cold water that occurs at this time of year near Point Año Nuevo, northwest of Monterey Bay.

But the project has other objectives. It may lead to the development of robot fleets that forecast ocean conditions and better protect endangered marine animals, track oil spills, and guide military operations at sea.

The undersea robots in Monterey Bay are part of a program known as Adaptive Sampling and Prediction (ASAP), funded by the Office of Naval Research. Its two co-leaders are Naomi Ehrich Leonard of Princeton University and Steven Ramp of the Naval Postgraduate School.

During the experiment, the ASAP system will determine what paths the underwater robots should follow to take the most information-rich samples, or measurements, of ocean activity. As the ocean changes, automated computer programs will update the sampling strategy under the supervision of the ASAP team. Most of the scientists will not be on site during the actual field experiment. The researchers will gather online in a virtual control room to share observations and make important decisions about necessary changes to the field operation as it is under way.

 Slocum Gliders
David Fratantoni of the Woods Hole Oceanographic Institution and Naomi Ehrich Leonard of Princeton University prepare Slocum gliders for launch into Monterey Bay.

The underwater robots, known as gliders, will take the ocean's temperature, measure its salinity, estimate the currents, and track the upwelling. Two types of gliders will be deployed: Spray gliders and Slocum gliders. The Slocum gliders belong to David Fratantoni of the Woods Hole Oceanographic Institution; the Spray gliders belong to Russ Davis of the Scripps Institution of Oceanography.

In contrast to typical ocean-observing systems, which are static, the mobility of the gliders allows them to capture the dynamic nature of the ocean, which is always shifting in time and space.

Inspired by the behavior of schools of fish, Leonard's group at Princeton created algorithms that allow the gliders to self-choreograph their movements in a series of rectangular patterns. The patterns span a large area the scientists mapped in Monterey Bay.

On a day-to-day basis, the control algorithms allow the gliders to make decisions independently about how to alter their course, without any input from humans. This day-to-day autonomy enables the gliders to move according to the organized patterns, even as they fight strong currents.

As the ocean changes and the robots detect new features, the ASAP team will reorganize the patterns to help guide the gliders toward ocean areas of interest such as eddies and thermal fronts. This process, called "adaptive sampling," should dramatically improve knowledge of the ocean and the ability to predict its chaotic behavior.

For related information, go to www.isa.org/sensors.