Maritime intelligence platform Eyesea and EVI Safety Technologies have unveiled an artificial intelligence–driven system that can automatically detect and report containers lost overboard, using a ship’s existing onboard cameras. After development and validation in simulated environments, the technology is now moving into trials on operational vessels.
The system applies computer vision and machine-learning algorithms to standard deck and bridge cameras, enabling it to identify containers as they fall from a vessel or when they are sighted drifting at sea. Once an incident is detected, the platform automatically records the event and transmits structured, geo-tagged data to registered shipowners, flag administrations and, where relevant, coastal authorities.
The rollout comes as the International Maritime Organization prepares to introduce mandatory reporting rules for container losses. These regulations will require vessels to promptly report the location, number of containers lost and any available details on their contents, particularly if dangerous goods are involved. Eyesea and EVI say their system is designed to help crews meet these obligations by generating immediate, standardised reports supported by visual evidence.
By automating detection and notification, the technology aims to reduce delays and inconsistencies that often accompany manual reporting. It also addresses a persistent safety issue: many container losses occur without being witnessed on board, or are reported only after significant time has passed, limiting the ability to warn nearby traffic and authorities.
Commenting on the launch, EVI Safety Technologies chief executive Vladimir Ponomarfev said the application of AI could finally bring meaningful progress to a long-standing maritime problem. “It’s amazing to be able to apply advances in machine learning and AI to help the maritime sector tackle an environmental and safety challenge that has been underserved for decades,” he said.
Container losses remain a recurring feature of global shipping incidents, with multiple cases reported in recent weeks, underlining the industry’s push for faster detection, better data and more transparent reporting.





