Thwarting Terrorism: Projects improve search and communication in ocean depths
Dangerous water conditions, hidden shipwrecks, enemy submarines—all manners of threats can lurk within the ocean’s depths. Princeton professors Naomi Leonard ’85 and Sanjeev Kulkarni are currently working on two separate projects, each funded by the Office of Naval Research, to address these concerns and enhance security at sea.
Leonard, a professor of mechanical and aerospace engineering, is co-leading the multi-institutional Adaptive Sampling and Prediction project. The project, which unites researchers at nine institutions and recently received an additional two years of funding, is developing ways to gather information about large areas of the ocean using underwater robots that coordinate their movements without human guidance.
Equipped with sensors and controlled by computer systems developed and based at Princeton, a fleet of six robots spent the month of August 2006 roaming a 300-square-mile area in Monterey Bay, Calif., as part of a large experiment. By surfacing regularly to send data back to a server and receive instructions from the control center, the robots worked with one another autonomously to ensure broad coverage of the vast expanse.
“These robots will provide a way to assess fairly rapidly whether an area is safe,” said Leonard. For example, she said, robots able to measure visibility and strength of currents could one day be used to determine when to deploy relief boats after a tsunami. “They allow you to make decisions without being overly conservative.”
During the California experiment, the robots gathered information about salinity, temperature and currents, documenting an upwelling of cold water. In the future, similar control systems and robots could use light-detecting sensors to detect submarines, which are often naturally coated with bioluminescent organisms. Other robots may one day use sonar to track the movements of boats.
Their ability to do so may be enhanced by the work now being done by Kulkarni, a professor of electrical engineering. In collaboration with researchers at Virginia Polytechnic Institute and State University, he is developing ways to process sonar data more intelligently and efficiently than is currently possible using pattern recognition techniques.
“The data being collected by cameras are absolutely overwhelming, so we need to use computers to pare down the information,” Schwartz said. “Oftentimes, the final decision as to whether something is a threat will be made by a human, but the computer systems will help them know where they should be looking.”