AI and Continuous Operational Data: A New Frontier for Maritime Safety Culture

## AI and Continuous Operational Data: A New Frontier for Maritime Safety Culture A growing debate within the maritime industry is sharpening focus on what artificial intelligence can — and cannot — tell us about safety culture aboard commercial vessels, with technology advocates and sceptics alike weighing in on the practical value of AI-driven monitoring systems. Writing for Splash247, Osher Perry, founder and CEO of ShipIn Systems, argues that continuous operational data collected through AI camera systems has the potential to deliver meaningful, actionable intelligence on safety, risk, and crew performance at sea. His commentary responds directly to an earlier piece by Alex Byelyavtsev, which questioned whether an AI camera is capable of genuinely measuring safety culture — a distinction Perry acknowledges as important but ultimately surmountable. 'The question is not simply whether a camera can see safety culture, but whether the data it generates over time can reveal patterns that human observation alone would miss,' Perry contends. The argument centres on the cumulative value of continuous monitoring: where a single snapshot may offer limited insight, longitudinal operational data can surface behavioural trends, near-miss frequencies, and procedural deviations that inform a more nuanced picture of onboard safety. The discussion arrives at a pivotal moment for the maritime sector. Shipping companies are under increasing pressure from regulators, insurers, and charterers to demonstrate robust safety management systems, while crew welfare and human error remain leading contributors to incidents at sea. AI-assisted monitoring is being positioned by some technology providers as a tool to bridge the gap between paper-based safety management and real-world operational behaviour. Critics, however, raise legitimate concerns around crew privacy, the risk of data misinterpretation, and whether algorithmic systems can adequately account for the complex human and cultural dynamics that define safety culture on a working vessel. Perry's response acknowledges these tensions, framing the technology not as a replacement for human judgement but as a supplementary layer of intelligence for fleet managers and safety officers. As AI adoption accelerates across port operations, voyage optimisation, and predictive maintenance, the debate over its application to crew safety monitoring is likely to intensify. Industry bodies, flag states, and classification societies are increasingly being called upon to provide guidance on the ethical and operational frameworks within which such technologies should be deployed. The full commentary by Osher Perry is available via Splash247.