A new computer algorithm could speed up the process of finding people lost at sea, potentially saving thousands of lives, researchers have said.
Every year, thousands of people die at sea in ship and plane accidents, and finding them rapidly is crucial as survival chances plummet after six hours.
The algorithm combines movement patterns with data about ocean currents, making it far faster to track down objects or people lost in the sea.
Read more: Man lost at sea survives because of his jeans
Professor George Haller, of the Swiss Federal Institute of Technology, Zurich (ETH), said: “Our hope is this method will become a standard part of the toolkit of coast guards everywhere.”
The researchers identified paths along which objects on the surface of the sea tend to float – and checked their theories using U.S. Coast Guard data.
They found that buoys and manikins thrown into the water gathered along predictable curves in the water.
Objects tend to move along predictable curves (ETH)
Lead author Dr Mattia Serra, now at Harvard University, Boston, said: “Our work has a clear potential to save lives.
“Our results are rapidly obtained, easy to interpret and cheap to implement.”
Read more: Indonesian teen lost at sea for 49 days ‘prayed every day’
Using sea-surface data, Dr Serra and colleagues identified a phenomenon they dubbed TRAPs (TRansient Attracting Profiles).
Mathematical results showed these are a few “special curves”, or short-term trajectories, along which objects on the surface will float.
Invisible to the naked eye, they can be extracted and tracked from instantaneous ocean data using the researchers’ methods.
This enables quick and precise planning of search paths.
Current operations use models based on sea dynamics, weather prediction and on site observations to produce a missing person’s probability map”.
But these are notoriously unreliable owing to changing tides and coastal currents, along with challenging conditions.
Haller said: “Of several competing approaches tested in this project, this was the only algorithm that consistently worked in situ.”
