Orca AI’s Autonomous Cargo Ship Successfully Navigates 500 Miles Without Collision

First of its kind in the field of fully autonomous commercial cargo vessels, Orca AI's integrated cockpit on the Suzaka proved to be more than successful at self-navigating the remarkably busy Tokyo Bay. Traveling 500 miles at a supposed 99% autonomy, the 750-ton vessel avoided all obstacles in its path all on its own.

The successful journey is a powerful moment for Orca AI and its sophisticated software, proving that the Israel-based company has a form of technology that could rewrite the future of the commercial maritime industry. The company was first founded in 2018 and has since made it its mission to evolve the way waterways are navigated through the use of powerful cameras and sensors in line with intelligent AI software.

In Sept. of 2021, Orca AI partnered up with Japanese shipping company Nippon Yūsen Kabushiki Kaisha, or NYK Line, to evaluate and assess the potential of the firm's software, called Automatic Ship Target Recognition System. NYK Line's Suzaka was chosen by yet another player in the field of self-driving vessels, titled Designing the Future of Full Autonomous Ships (DFFAS).

Retrofitted with Orca AI's aforementioned software, Suzaka traveled a total of 790 km (491 miles) starting from Tokyo Bay. The journey took about 40 hours to complete and, according to Orca AI, required next to zero human intervention in its adventure, slipping past several tug boats and other assorted vessels in its path to Tsumatsusaka port in the Isle Bay.

Although but a mere trial run, the maiden voyage for Orca AI's Automatic Ship Target Recognition System proved immensely worthwhile, as Tokyo Bay boasts among the most congested waterways in the world. Suzaka, via the help of the AI software and without any aid from humans, made a grand total of 107 collision avoidance maneuvers and avoided between 400-500 assorted ships merely on its outbound voyage.

Orca AI's system sits mainly within the cockpit as a command center-like setup aided via 18 different onboard cameras that utilize a host of different procedures, main among them being a 360-degree view, range estimation, tracking, real-time detection, classification, and more. This data is all collated into the software through deep learning, which Orca AI terms the "human watchkeeper," despite its digital nature.

"We are honored to collaborate with the DFFAS consortium led by NYK group to drive automation and autonomous capabilities in commercial ships in some of the most congested waters in the world," said Yarden Gross, CEO and co-founder of Orca AI. "The world's first commercial autonomous voyage is a significant milestone in this journey and we expect to see big shipping companies implementing advanced AI and computer vision technologies to materialize the autonomous shipping vision."

The journey would not have been possible without nearly a year's worth of collated data made via Suzaka and Orca AI's software. These initial test runs proved to allow the system to learn the varied nature of Japan's waterways and shorelines, with one such simulation conducted between Feb. 26th and Mar. 1st proving to be successful.

"With the great majority of Japan's international trade relying on shipping, it's been imperative for us to find technological solutions that can make shipping safer. We thank Orca AI for helping us find the ultimate computer vision solution to enhance situational awareness on ships, despite the obstacles the global pandemic has presented," said Dr. Hideyuki Ando of NYK Group. "Their successful delivery of the entire project, thousands of kilometers away from Japan and with complex hardware and software being shipped and remotely updated regularly, is a prime example of the potential technology has to transform the industry."