Point Your Phone at the Surf and an AI Rip App Will Show You Where the Rip Might Be

Mandana Ghanavati reckons your phone could help before you swim. Snap the surf first. The app does the rest.

She’s a senior researcher at UNSW. Her team made RipEye. It’s an AI rip app. It marks where a rip might be on your photo. Surf lifesavers at some NSW beaches are trialling it this summer.

“We are hopeful that this technology can help people to spot the rip,” she told the ABC.

Can’t you just see them?

Not really. Ask Jaz Lawes. She’s the national research manager at Surf Life Saving Australia.

“Rip currents are notoriously difficult to identify and it’s really hard to teach people how to identify them as well.”

Can you spot a rip? Half of us reckon so. Only one in seven actually can. That’s from SLSA’s own research.

How does RipEye know?

It learned from thousands of beach photos. Experts had marked where the rips were in each one. Ghanavati’s team then trained the AI on those.

Is it always right? No. Eight times out of ten, the researchers say.

That’s still a lot better than us. A trained human eye gets it about half the time, UNSW reckons.

Years in the making

RipEye didn’t come out of nowhere. The boss of the project is Mitch Harley. He’s at UNSW’s Water Research Laboratory. It got Australian Research Council money back in 2023, with SLSA as a partner.

The original pitch was software that could pick rips from smartphone video. Toby Walsh is in on it too. He’s UNSW’s big AI name. So is beach scientist Rob Brander.

Shane Daw from SLSA didn’t mince words when it was announced.

“Rip currents are the number one beach hazard globally, responsible for one in five drowning deaths,” he said.

By September 2025 the team was getting close. Harley told UNSW they were about six months off testing it with SLSA groups.

“We’re about six months away from pilot testing this with various groups with SLSA.”

This summer’s trial is that test. It runs off a normal phone photo. Nothing fancy.

There’s a game in it too. Up comes a beach photo. You guess the rip. Then the AI shows you where it reckons the rip is. Yours gets compared with the machine’s.

Teaching people is the hard part, Lawes said. That’s what the game’s for.

Built for the people most at risk

RipEye works in 11 languages. Young blokes are one target. Visitors from overseas are another. So are people from culturally and linguistically diverse communities.

They’re the groups more likely to drown, Ghanavati said.

Some numbers. There were 66 beach drowning deaths between July 2025 and July 2026. Almost half were partly down to rips. That’s from the latest National Drowning Report, the ABC says.

Which beaches get it?

Not decided yet. SLSA is still working out which surf clubs get it. The trial runs over this summer.

Lawes reckons it could help train new lifesavers. Teach the public too. Where the flags go, too. Patrols could use it for that.

Country towns could get a go as well. SLSA NSW runs a Beach to Bush program for inland communities. The app could tag along.

Fewer deaths last summer?

Yes, a few. Fifty people drowned on the coast over summer 2025-26, SLSA counted. The summer before it was 63.

Rip incidents dropped too. Down 43 per cent on the five-year average.

Most of those deaths were nowhere near a patrol, though. Two in three happened more than a kilometre from a surf lifesaving service.

Not the first AI on the coast

SLSA has already tried AI cameras at rock fishing spots. Little Bay was one. That’s in Sydney’s south-east. Kiama, down on the South Coast, was the other.

Rips are the number one coastal drowning hazard in the country. That’s according to Lawes.