Lifeguards do play a vital role in the prevention of drownings, but at large busy beaches, they can’t always keep track of all the swimmers at all times. That’s where Sightbit is designed to come in, as it uses artificial intelligence to catch what lifeguards may miss.
Developed by a startup that originated at Israel’s Ben-Gurion University of the Negev, Sightbit incorporates multiple video cameras located along a stretch of beach. These transmit live video to a computer vision system located in a central lifeguard tower, which continuously analyzes the footage utilizing deep learning-based algorithms.
That system is able to simultaneously detect and track multiple people in each shot, using object recognition to identify things such as swimmers in distress, untended children, and watercraft that may be approaching too close.
Should any of these be detected, the system sounds an alarm in the tower, plus it zooms in on the swimmer, etc so that the lifeguard can get a better look on the computer screen. If they determine that it isn’t a false alarm, they can proceed to perform a rescue.
Sightbit will additionally draw attention to threats that don’t necessarily require an immediate response – such as potentially dangerous riptides – so that lifeguards know to keep a particularly close eye on them. That said, it can reportedly also be used in place of human lifeguards, automatically summoning rescue teams when trouble is detected.
The system is about to be trialled by the Israel Nature and Parks Authority, along the 5 mile-long (8-km) beach at Palmachim National Park. Pilot projects in Europe and the US may soon follow.
AI system designed to aid – or perhaps replace – human lifeguards [New Atlas]