The data foundation challenge: why optimisation fails without reliable data

Around 70% of vessels still rely on noon reporting, providing only a single daily snapshot of how a ship is performing. Even where high-frequency sensor data is available, miscalibration, missing values and weak integration plague the chain. This insight examines why robust data is the prerequisite for any meaningful optimisation, and why companies who skip this step risk building automation on unstable foundations.

Before looking into how operational optimisation is applied in practice, it is important to consider the quality of the data that supports it. This sort of dynamic optimisation relies on reliable operational data, including speed, fuel consumption, draft, weather conditions, and engine parameters.

These data points need to be captured accurately and often enough to support analysis. In reality, however, data environments across global fleets remain uneven, with around 70% of vessels still reliant on noon reporting, where key operational data is recorded only once a day.

Even where operators invest in performance tools, the quality of the underlying data is a constant vulnerability. High-frequency data captured from sensors only helps if it is complete and consistent, and if crews and shore teams can rely on it. Otherwise, instead of optimising, teams end up spending their time just trying to figure out what data they can actually trust.

Penny Haire, CEO of Tidetech, discussed how a key part of this challenge is in separating true vessel performance from environmental influence. She explained how “the only way to really understand the baseline performance of your vessel is to get rid of the noise.”

As Dr Timoleon Plessas, Founder, Sealion Engineering, explained, “you need reliable data, not just from a few sensors, but from every sensor in the system. It’s a chain of interdependent links, if even one fails to deliver, the integrity of the entire chain is compromised. After all, you cannot optimise what you cannot measure, that’s a fundamental law.”

Reliance on noon reporting has long been standard practice in the maritime sector, but it only provides a limited view of how a vessel performs throughout a voyage. Research has shown that relying on noon report data can cause uncertainty when evaluating vessel performance, as a single daily average detracts from the variations that occur during the day.

Despite the benefits of modern digital tools, even with more advanced monitoring systems in place, data quality issues continue to plague crews. This can be seen in reviews of ship performance data processing, which underline common problems like miscalibrated sensors, inconsistent timestamps, missing values, and poor integration between onboard systems.

This challenge is increasingly being addressed through more structured data environments and validation layers, including platforms such as DeepSea Technologies’ Vessel Core, which aim to improve consistency and reliability across onboard systems.

These issues can affect the accuracy of optimisation models and the confidence operators place in them. When optimisation results fail to meet expectations, the tools are often blamed, even though the problem often originates elsewhere.

For example, if the data input into a system is incomplete, inconsistent, or unreliable, even the most sophisticated models will struggle to produce dependable results.

There have been recent efforts to improve data infrastructure, with industry standards like ISO 19847, which sets out the requirements for shipboard data servers used in machinery data collection and aims to create more consistent and reliable data environments onboard vessels.

However, effective adoption remains very limited across the industry, and the underlying issue is that data collection and validation processes remain weak.

Without reliable high-frequency data, optimisation systems struggle to deliver accurate outputs or build the trust required for consistent use.

Strengthening data foundations is a critical first step in closing the gap between optimisation insights and operational performance.

Read more in our latest report here.