Underwater Robots Know Where They're Going
Twisty underwater ravines and seas with moving icebergs provide tricky terrain for an autonomous underwater vehicle (AUV) to navigate; that’s why underwater vehicles sent to investigate such areas rely on remote piloting from shipboard.
That won’t be necessary for much longer, according to engineers from Stanford University’s Department of Aeronautics and Astronautics and the Monterey Bay Aquarium Research Institute (MBARI). They have developed a system that allows an AUV to consider both an existing terrain map and its own view of obstacles in its path as it “flies” over the sea floor.
Stanford doctoral student Sarah Houts took an existing terrain-relative navigation system developed by Stanford and MBARI—one that allows an AUV to navigate by matching its altitude to a terrain map—and added algorithms that enable the vehicle to plan its route to steer around obstacles spotted ahead. Eventually, Houts indicated, the system should be able to go into unmapped areas and find its way around safely.
Houts hopes to adapt this technology to an upcoming MBARI effort to use AUVs to follow icebergs around and sample them. NASA’s program on Astrobiology Science, and Technology for Exploring Planets might someday use a similar system to monitor asteroids. (NASA is funding Houts’ work.)
The system had its first test in Monterey Bay earlier this month, flying over an underwater cliff at a constant altitude. It will undergo trickier tests before the end of the year, and Houts expects it to be fully operational next year.
Photo: Earlier this month, this autonomous underwater vehicle tested software that enables it to adjust its path according to obstacles or uneven terrain spotted in its path. Photo credit: Sarah Houts/Stanford University