LONG before Starlink became a household name and a critical technology connecting seafarers on the high seas, Orca AI founders Yarden Gross and Dor Raviv spotted an opportunity on the horizon.

Cloud computing was proliferating across industries and sectors, and the two entrepreneurs, both experienced at sea and in tech, wanted to leverage the technology to improve ships’ safety and efficiency.

“We wanted to create a platform that could help crews in real time by reducing workload and increasing safety, while simultaneously leveraging data to show trends and create actionable insights for fleet managers,” Gross told Lloyd’s List in an interview.

Orca AI’s automated watchkeeper, the SeaPod, uses a combination of cameras installed on a ship’s mast and onboard sensors to capture its surroundings in real time, and artificial intelligence to deliver actionable insights to the ship’s master and shoreside management. The newest model released in January is equipped with cameras providing a 225-degree field of view, while its thermal vision cameras provide 100-degree FOV. 

The images are live-streamed to a monitor at the bridge, while its computer vision algorithms are trained to detect dangers and alert crews to such risks. Meanwhile, the SeaPod’s thermal vision lets masters understand their surroundings when sailing through congested waters in darkness.

“Crews face heavy workloads and fatigue, and this makes staying alert and focused more complex, especially when navigating through congested waters at night,” Gross said.

For Gross and Raviv, both experienced in the challenges of navigating congested waters at night, watchkeeping was the obvious task to start automating, Gross explained.

“We wanted to not only automate the lookout’s role and improve it with capabilities like thermal vision, but also to leverage our data and AI to create an early warning system that would identify threats.”

Moreover, he said, SeaPod’s extra pairs of eyes means fewer lookouts are needed on the bridge, freeing them up to rest or perform other tasks, and reducing the overall workload.

Since its first commercial launch in early 2021, Seapod has been installed on more than 700 ships, gracing the fleets of major owners such as NYK, Marubeni and Maran Tankers. Gross expects that figure to more than double this year.

Some customers have seen close encounters and near-misses drop by 55% in six months, Gross said. Meanwhile the company recently stated that a survey it carried out on 267 vessels using the SeaPod found a 172,716 tonne reduction in CO2 emissions, despite a 26% increase in minimum distance travelled. This was largely due to a reduction in sharp manoeuvres and speed drops, it said.

For Gross, automating watchkeeping is only the beginning, and Orca AI’s eyes — human and digital — are set on autonomous shipping.

In May 2022, the company, along with partners NYK, DFFAS and others, completed a 40-hour voyage that was 99% autonomous on board the 749 gt Suzaku (IMO: 9853357) from Tokyo Bay to the port of Tsumatsusaka.

Gross envisions shipping becoming more like aviation, where captains fly the planes but most of the work is automated.

“An ocean-going commercial vessel will always have a crew, but you will be able to reduce it and its workload while increasing operational efficiency,” he said.

Recent events in the Red Sea present another use case for an AI-powered system that can use computer vision to interpret images in real time and provide early threat detection. Gross said Orca recently signed an agreement with an undisclosed navy, and more navies have enquired since.

“Our system can offer complementary threat detection to existing measures navy ships use, like radars” he said. “It can help plug in certain gaps.”

‘A captain’s skills cannot be learnt by an algorithm’

For Capt Mark Ervin, the master of the Marubeni-owned supramax Crimson Delight (IMO: 9732163), SeaPod is a much welcome addition to his ship.

Speaking to Lloyd’s List on board the Panama-flagged bulker on a snowy January day in the port of Camden, New Jersey, Ervin said the system is particularly helpful in congested waters during nighttime, when visibility is limited, and the radar is crowded with signals.

Rather than seeing dots on the horizon, the thermal vision cameras allow him to see the ships’ design and features on a monitor.

Ervin said it also allows for more efficient use of the ship’s most important resource — its seafarers. 

“The biggest challenge is navigating during limited visibility in congested areas, which requires two more additional lookouts. But with Orca AI we can reduce the number of additional crew,” he said.

Ervin believes the system might lead to a reduction in safe manning requirements in the future, but he is not worried that it will make seafarers obsolete.

“A captain’s skills cannot be learnt by an algorithm; a computer only relies on what you feed it,” he said.

“It doesn’t have the ability to adapt and change to a situation like a human. The techniques and tricks cannot be carried out like a human.”