An Israeli maritime startup is hoping to prevent ship collisions using artificial intelligence (AI) and data analytics.

Orca AI is working on a possible solution to the issue of ship collisions, 90% of which happen in and around crowded ports and docks. With an estimated total cost of $20 billion, ship collisions are occurring with greater frequency thanks to overcrowding and larger ship sizes.

“When you get close to a port, it is usually very congested from all sorts of vessels like container ships and small shipping vessels. The existing navigation tools like radar and GPS are not sufficient for such waterways,” said Yarden Gross, co-founder and CEO of Orca AI. “In such a situation, the captain and crew start aligning the vessel based on their instinct. But then, bad weather conditions also need to be considered. For instance, when there is a lot of fog or haze and/or its night time, you have low visibility. Eventually, it doesn’t matter how much experience you have, you are prone to making mistakes, and that can be catastrophic.”

To combat the issue, Orca AI is building a system that relies on thermal cameras with built-in vision sensors. The thermal cameras are combined with low-light, high-definition cameras attached to radar and GPS sensors, enabling ships to navigate through obstacles and avoiding potential collisions. The system, according to Orca AI, learns about the maritime environment by analyzing the data collected by the sensors and through the application of AI, eventually adding to the system’s autonomous capabilities.

“At Orca AI, we are taking a piece of the autonomous technology puzzle and solving it. We want to make sure that ships can travel from point A to point B in the best and safest way possible. As more and more ships install our system, it will be possible to statistically understand the extent to which we reduce the collision numbers,” said Gross. “Right now, we are working closely with our customers to improve the system, the user interface and the overall capabilities of our software.”