AI at sea: Inside Zelim’s ZOE safety system

Sam Mayall, CEO of Zelim

When cruise passengers imagine a relaxing voyage, they rarely think about the technology quietly working behind the scenes to keep them safe. But for Sam Mayall and Doug Lothian, the founders of Edinburgh, Scotland-based safety innovator Zelim, ensuring passenger safety is a full-time mission—one that’s rapidly evolving thanks to the company’s AI-powered man-overboard detection and tracking system, ZOE.

Recently installed aboard the cruise ship Ambition, ZOE faced its toughest test yet. For Zelim, adapting the system for use in the complex environment of a cruise ship presented both technical and operational hurdles.

“Technically, this was the first time we deployed so many cameras in a single environment,” said Mayall. While the software itself had been proven, the hardware required rigorous load and soak testing to ensure it could handle real-time processing of multiple video streams without failure. “We needed to be absolutely certain the system would remain robust under constant pressure,” he added.

The operational challenges were no less significant. Installation had to be done while the vessel was in service—without disrupting the guest experience. “We had crews flying around the world to meet Ambition in different ports, including in the Caribbean,” Mayall said. “It was a logistical headache, but worth it.”

ZOE’s capabilities go far beyond just recording footage. In man-overboard scenarios, the system uses a two-step process: it first detects the fall, then uses pan-tilt-zoom cameras to track the person in the water—even in rough seas or poor visibility. “We use predictive tracking,” explained Mayall, “so even when the person isn’t visible due to wave action, the system anticipates where they’re likely to be, with almost 100% accuracy.”

The independent testing aboard Ambition, including evaluations by Lloyd’s Register, proved the system’s accuracy. Zelim even tested ZOE’s response to adult, child, and infant mannequins—going beyond ISO standards—and found it could accurately detect smaller bodies, treating them as distant adult figures.

ZOE is also the foundation for Watchkeeper, a forthcoming product designed to provide 360-degree awareness around the vessel, detecting not only people overboard but also navigational hazards like flotsam, small craft, and low-radar-signature threats. “Radar is great, but it struggles with small or low-profile objects,” Lothian explained. “Watchkeeper fills that gap, offering constant vigilance in all directions.”

The goal, Lothian says, is not to replace human watchkeepers, but to reduce their stress and workload. “Watchkeeper never gets tired, never gets distracted. It’s your best officer of the watch—24/7.”

With the maritime industry seeing an uptick in overboard incidents and increasing concern over bridge team fatigue, ZOE’s real-time AI capabilities may prove crucial—not just for cruise ships, but for tugboats, tankers, offshore vessels, and beyond.

“We’ve had great reception across the cruise sector,” Mayall said. “And we’re beginning to hear from vessel operators we hadn’t even approached. The interest is global—and growing.”

ZOE and Watchkeeper represent more than just the next step in maritime surveillance. They’re a leap toward proactive, AI-driven safety systems that not only save time and money—but lives.