Introduction
Cast your mind back only a few years ago, and the idea of collecting real-time vessel data seemed more like a dream than a reality. With the rapid evolution of technology, particularly Starlink, ships have been brought online at sea, allowing them to continuously gather, analyse, store and transmit data back to shore. What’s more, this approach is quickly becoming the new way to measure and select risk. Instead of using static, backwards-looking assessments, we’ve entered a new era of dynamic, data-driven underwriting and insurance.
The limits of traditional hull & machinery underwriting
To understand how Hull & Machinery insurance and underwriting has taken shape over the last few years, we need to look at what it was like before.
Hull & Machinery insurance has traditionally relied on periodic surveys, scheduled maintenance checks, and manual reporting from the vessel and its operators. Insurers assessed risk primarily using physical inspection reports, historical loss records, vessel age, class status, and regulatory compliance, rather than continuous operational performance. This approach creates significant gaps in information, fails to capture emerging risks and is an approach that’s no longer fit for purpose.
Let's look at Planned Maintenance Systems (PMS) - while they provide a structured record of routine inspections and servicing (mandated by flag states and executed by the onboard engineering team) they only show compliance on paper. The critical period between planned tasks and reactive repairs is largely invisible to insurers, meaning actual machinery condition and maintenance schedules often remain unknown.
To highlight this, we’ll use a mid-sized cargo vessel as an example. Its PMS is configured in accordance with OEM recommended inspection intervals. Each inspection is completed and recorded as expected, however between inspections, a cylinder exhaust begins to deteriorate. Four days later, the engine suffers a partial failure, leaving the vessel without power and delaying a charter.
Had real-time data been available, predictive maintenance could have flagged the early signs of wear on the cylinder exhaust, enabling the crew to intervene and prevent the costly downtime and repairs. Instead, the deterioration goes undetected between scheduled inspections and only becomes visible once it results in a loss. This is the gap that traditional approaches leave exposed – and this is the problem. Marine insurance remains largely static, backward looking and doesn't have the transparency required to price and structure suitable cover for high value assets. Isn’t it time we changed how we assess risks?
Can AI & predictive analytics redefine risk assessments?
AI, machine learning, and The Internet of Things (IoT) are giving Hull & Machinery underwriting a new level of insight and precision.Predictive maintenance platformsuse real-time sensor data from engines, generators and critical machinery to flag early signs of wear or stress, often long before they turn into failures and costly claims.Research indicatesthat predictive maintenance reduces fleet downtime by 50%, lowers maintenance costs by 40%, and decreases equipment failure rates by 60%. Across fleets and voyages, machine learning models can spot patterns and estimate the likelihood of machinery failure. Insurers can price risk more accurately and shipowners can plan maintenance proactively, moving underwriting from reactive guesswork to data-driven, precise decision-making.
To highlight how powerful predictive maintenance can be, let’s look at a recent example. Ceto worked with a client who had a container ship with bunkered fuel that initially tested withinISO 8217 standards.About 10 days later, the engine failed, leading to a $1 million insurance claim and a number of days off-hire. Only when the data was reviewed retrospectively did the telltale warning signs become apparent. Had predictive analytics and intelligence been in place, the failure could have been prevented, avoiding the insurance claim altogether. Insurers can price risk more accurately and shipowners can plan maintenance proactively, moving underwriting from reactive guesswork to data-driven, precise decision-making.
It's not just a buzzword, predictive maintenance is becoming the new standard for operational decision making. By tracking engine temperature variance, turbocharger efficiency, hull stress sensors and more, across voyages and vessels, machine learning models can predict component failures days or weeks in advance, turning reactive maintenance into proactive planning, saving time and money.
By tracking these metrics across voyages and vessels, machine learning models can predict likely component failures days or weeks in advance, turning reactive maintenance into proactive planning, saving time and money.
In a recent simulation, Ceto'smachine learningalgorithms detected a single abnormal main engine exhaust temperature reading on an LPG tanker. The crew were alerted with the changing severity over the course of four days, giving them clear, timely warnings that enabled them to safely intervene before an auto slow down and subsequent shutdown stranded the vessel and crew in the Red Sea. Had the issue progressed to failure, even a routine corrective action such as replacing an exhaust gas valve or cylinder head could have resulted in approximately 1.5 days of total off-hire once operational disruption, restart procedures and schedule recovery are considered. At $36,000 in off-hire, plus $7,500 in daily OPEX and $9,600 in off-hire bunker costs, the financial implications would have been $56,850. By identifying the anomaly early, the crew would have been able to take corrective action before safety, schedule and commercial performance were impacted.
From black box to glass box: The rise of “transparent vessels”
Think of modern solutions like Watchkeeperas giving vessels a “glass box” to see what’s happening onboard in real time. Engine performance, fuel use, hull stress; all the data that used to sit in logbooks or in delayed reports is now visible and importantly, actionable.
For owners and managers, this kind of transparency is a game-changer - they can spot issues before they become problems, and plan maintenance in a smarter way, enabling more control of their fleet. At the same time, insurers gain confidence by seeing what’s really happening, rather than relying on snapshots and historical claims. This means Hull & Machinery underwriting isn’t about guessing what might go wrong - it’s about making decisions based on what is actually going on.
How can data be used to negotiate terms with insurers?
This is a question we’re often asked at Ceto! The answer is two-fold. Predictive maintenance doesn’t automatically give you lower premiums, but it makes operations more efficient and reduces the likelihood of claims, which insurers notice and reward. It also gives brokers powerful evidence to negotiate with insurers.
For instance, a fleet using real-time monitoring may report:
• 20% reduction in minor machinery claims over 12 months
• 30% fewer unplanned downtime days
• Consistently lower engine temperature variance compared to fleet averages
While Maritime is relatively new in its adoption of predictive maintenance, across other industries, such as aviation, it’s well established, and the results demonstrate how powerful it can be. In mid-2024, easyJet reported that Skywise predictive analytics helpedavoid nearly 79 flight cancellationsover two months by identifying issues early and addressing them before they caused service disruptions. The likelihood of claims associated with disruptions, operational costs, or emergency repairs reduced, a key consideration for underwriters when pricing risk and negotiating terms.
Verifiable vessel performance data allows brokers to demonstrate that their client operates a low-risk fleet. In essence, vessels that are well-run, well-maintained, and proactive get recognised for “good behaviour”, and that can translate into more tailored, risk-reflective terms.
This is what’s meant by data-backed insurability: instead of relying on old loss records or periodic inspections, underwriters can see what’s really happening on the vessel day-to-day. The upshot? Shipowners get rewarded for taking care of their fleet, and insurers can price risk with a lot more confidence. Efficiency plus predictability is where the real operational upside comes in, because smoother, more reliable operations ultimately drive profitability.
What does the future hold?
More and more connectivity! As vessels come online, Hull & Machinery insurance is shifting from a reactive product to a proactive partnership. Shipowners who share real-time operational insight gain smoother operations and more tailored terms while underwriters gain clarity and confidence in risk. The result is a smarter, fairer model and one that rewards transparency, performance and collaboration.
