Using AI to enhance navigational safety

Using AI to enhance navigational safety
Orca uses AI to identify ships and other objects that are not always visible to the human eye

In the last five years approximately 4000[1] collisions and incidents have been reported annually, with 75-96 per cent of these marine incidents attributed to human error[2].

Speaking during Digital Ship’s webinar on how AI can be used to enhance navigational safety and save human lives, Philip Nielsen, general manager Europe of maritime start-up Orca AI explained that the increase in global marine traffic and larger size of ships sailing today have contributed to heavier ocean congestion. Furthermore, Nielsen believes that while officers today are more experienced when it comes to operating technology, there is greater uncertainty on how to use navigation equipment for safer decision making.

Founded in 2018, Israel headquartered start-up Orca AI has developed an artificially intelligent navigation solution that provides a full and accurate picture of the navigational environment to boost human capabilities and reduce accidents at sea.

The solution comprises a lookout unit with computer vision, which provides a picture via a large screen to the bridge. It uses three high-definition cameras and three thermal cameras that can detect small objects early in low visibility situations. The visual field created by the cameras is integrated with another nine navigational sensors that already exist on the bridge, such as radar, which is fed into a learning algorithm that has the capability to analyse all the data. This provides crew with greater clarity on difficult and dangerous situations to help them make safer decisions.

Orca’s tool is able to identify between different ship types and objects. “This is super important because when you provide an insight to a captain, you have a very big difference if you have a boat crossing from a fishing vessel or a large container ship,” Nielsen explained.

Orca’s thermal camera identifies all targets in low visibility conditions

A sophisticated alarm provides real-time alerts to the crew and captain. Highly dangerous situations are prioritised, while those requiring less urgency generate a less intrusive alert, reducing alarm fatigue for the crew.

The data used by Orca is collected and aggregated onboard. Data is kept on board a vessel for at least two months and only relevant information is uploaded to a private cloud. Thanks to the use of AI, the most relevant data according to a shipping company’s Safety Management System is cherry-picked and displayed on the dashboard. Clever algorithms learn from the data to improve their function over time.

During his presentation, Nielsen also brought to light an issue he often hears shipping companies speak about – lack of effective onboard training for offices based on real case scenarios. To target this, Orca provides users with a high-quality touchscreen tablet which replicates the bridge display, enabling the captain from any location to view what is happening on the bridge. The tablet benefits from a record function, which can be used to record officers’ responses in different scenarios and played back to them to train for future situations.

Orca uses AI to enhance situational awareness, making non-visible objects visible

Integration and installation

Orca AI does not aim to compete or replace radar, instead it is a navigation aid that adds an extra level of security. According to Nielsen, all components included with Orca, both the hardware and the software, has been specifically built for the marine environment. This means it is straightforward to integrate into other existing navigation sensors. Orca is also integrating with all major OEMs in the market and developing its own National Marine Electronics Association (NMEA) approved standard to export information from the cameras and the vision sensors to standard equipment or the same Ethernet protocols as other equipment works on. In the future Orca will aim to integrate with other navigation equipment onboard like ECDIS.

Orca can be retrofitted to any vessel, typically within less than 6 hours.

In April this year, Orca announced a $13 million in Series A funding, taking its total raised to over $15.5 million. This will be used to support the company’s investment in technology, international expansion and growth.

To find out more about Orca AI, watch our webinar here.

[1] EMSA Annual overview of marine casualties and incidents 2020

[2] Allianz Global Corporate & Specialty; Safety and shipping review 2020