- Technology company Sea.AI helps ships avoid collisions and unseen hazards at sea.
- The company's software uses thermal cameras and AI to identify underwater objects and alert mariners.
- This article is “Build IT” is a series that highlights the digital technology trends that are disrupting industries.
Will Harris has been a professional ocean racer for over 10 years, skippering and sailing with Team Malizia and spending more time away from land than the average person.
“What's so difficult about being at sea is when you're in the middle of the ocean, you're really one of the most isolated people at that moment,” he said.
Safety is a top priority for mariners. For decades, ships have relied on technology such as radar to detect large obstacles. But objects like trash, wreckage, fishing boats and buoys often remain invisible until a collision is imminent.
To avoid these obstacles during the race, Harris uses software from Sea.AI, a startup that uses artificial intelligence detection and proprietary thermal cameras to give boats a better view of the ocean. Founded in 2019, the French company has collected data on more than 10 million objects that can't be detected by radar. Soren Guero, Sea.AI's head of marketing and communications, told Business Insider that 500 boats use the company's AI technology.
Sea.AI's most widespread use is in boat races to avoid collisions: “These are multi-million pound boats,” Harris said. If a collision were to occur, he added, the race would have to be stopped because it would be “obviously very dangerous.”
Harris began using Sea.AI the same year it was launched and has sailed more than 100,000 miles using the technology. “For us, it’s a technology race, so we’re really trying to use data to our advantage,” he said, adding that the ship is equipped with several sensors that measure metrics such as speed, depth, fuel and the ship’s infrastructure.
The boat he co-skippers with Boris Hermann, Malizia Sea Explorer, is powered by Sea.AI and helped Team Malizia take second place in the first leg of the Transat CIC race from France to New York on May 6th.
How the technology works
Sea.AI founder Rafael Biancaré previously worked in the self-driving car industry. When he got his first boat, he noticed a gap in the technology for surface vessels. When sailing at night, “we didn't have the right systems to help us see ahead of the boat,” Guéroux said.
Sea.AI uses large thermal cameras that can be mounted on the front of a ship or on a mast to scan the water's surface. The software uses thermal imaging and night vision capabilities to detect objects that are hard to see with the naked eye.
Images from the cameras can be accessed through an app or connected to the ship's multifunction display system, and if an object is detected, the software automatically sounds an alarm, giving the captain enough time to change course.
Sea.AI's system then records the object, training its software to detect more types of obstacles.Harris said Sea.AI's alarms are a “very important” feature because sailors aren't always looking at the cameras — “they're looking at the sails, they're sleeping, they're resting.”
Harris recalled that the team avoided a collision while sailing near Gibraltar at “25 to 30 knots” (about 29 to 35 miles per hour). He said he wasn't monitoring the cameras, but heard an alert from Sea.AI that a fishing boat was about 100 meters away. “We probably had 10 to 15 seconds before we hit it.”
Building a more comprehensive navigation system
Sea.AI was developed to be used in conjunction with radar technology and the Automatic Identification System, a tracking system that allows ships to locate each other's geographical locations.
The company has partnered with Garmin, a manufacturer of multifunction displays, to enable Sea.AI to integrate more seamlessly into Garmin software. Multifunction displays centralize information for mariners monitoring multiple systems that help operate a vessel.
“Floating object detection has always been an important area, so we were intrigued by the innovative technology the company has brought to market,” said Andrea D'Amato, business development manager, marine, Garmin International.
“Our display can be split into two windows,” D'Amato told BI, “so, for example, one window can focus on the radar and the other window can focus on the Sea.AI capabilities.”
Harris said the MFD has side-by-side displays for Sea.AI, the automatic identification system and radar. The radar and AIS displays help with navigation, while the Sea.AI display shows the location of objects relative to the ship's position. “You immediately know which direction you need to look to see the object,” he added.
Romain Attanasio, captain of the Fortinet Best Western, recently used Sea.AI to guide his ship across the Atlantic Ocean, and told BI that avoiding the many fishing boats along the coast was one of the biggest navigational challenges, especially when entering and leaving port.
Attanasio said many fishermen don't use AIS because “they don't want other fishermen to know where the fish are,” making it harder to spot their boats. He added that fishermen's nets are also an invisible hazard because they pack large amounts of fish and rope just below the surface of the water, which is largely invisible to AIS and radar technology but is easier to monitor using Sea.AI imagery.
Sea.AI, the Automatic Identification System and radar each have their advantages and disadvantages, Guérou said. “Each sensor has its own capabilities,” she said. “We are an optical sensor, so if there's really heavy rain, for example, it's hard for us to handle it.”
The rise of AI in shipping
Gouraud told BI that Sea.AI is expanding with the launch of Brain, a software-only machine vision product that can identify underwater objects. Unlike previous software that only worked with the company's camera units, Brain is compatible with third-party thermal cameras. The goal, Gouraud said, is to reduce the cost of using the Sea.AI system by eliminating the need to purchase both software and cameras from Sea.AI.
The company also plans to enter other markets, such as cruising, commercial shipping and ferries. Guéroux said Sea.AI recently conducted trials with search and rescue vessels in France, the UK and Germany.
Guero said the future for Sea.AI will be building technology with “auto-avoidance” capabilities that reduce the need for human intervention. Harris' ship uses a prototype of this feature, which connects to the ship's steering system to automatically avoid collisions with small objects nearby. “It's still in development, but you'll definitely see more and more developments in the future,” he said.
