AI Confronts Seagoing ‘Rogues’

Once dismissed as seafarers’ tall tales, monstrous waves that seemingly rise out of nowhere to tower over surrounding waters are now thought to be regular culprits in the sinking of large ships.

A new tool from UMD researchers could give mariners and offshore oil platform residents up to a five-minute warning of these “rogue waves” so they can batten down the hatches or position their ships to survive the wild ride.

The new system was developed by postdoctoral researcher Thomas Breunung and Distinguished University Professor Balakumar Balachandran, both from the Department of Mechanical Engineering, and presented in Scientific Reports.

To create the tool, they trained a neural network—an artificial intelligence system that mimics human cognition—to distinguish ocean waves that will be followed by a rogue wave from those that will not. The training data consisted of billions of regular waves and thousands of rogues, recorded in 14 million 30-minute-long samples of sea surface elevation measurements from Pacific Ocean buoys.

After training, the system correctly predicted the emergence of 75% of rogue waves one minute into the future and 73% of rogue waves five minutes into the future.

Adding water depths, wind speeds and buoy locations at sea into the data could further boost the system’s accuracy—and it could be applied far beyond ups and downs at sea, Balachandran says. “Our data-driven approach could be useful for understanding and predicting other extreme events associated with, for example, climate change.”