ERRORS in vessel operational data can arise from a number of sources and may not be spotted by onboard crew, says StormGeo product manager Selin Sari in this podcast. She outlines some of the consequences of these errors, such as incorrect emissions calculations and technical performance analysis, all of which are relied on when making commercial decisions.

Sari explains why both manually-recorded and high-frequency sensor data can be incorrectly logged and says that analysts need to be aware of this possibility when looking at datasets supplied to them. Based on StormGeo’s experience, she gives examples of situations when analysts have spotted abnormalities that had been overlooked onboard.

Even establishing what a ship’s “normal” performance looks like requires insight. The biggest external factor affecting performance is the weather, so StormGeo’s own data is a valuable resource to be able to take that into account, she suggests.

And although measurements can be taken of such things as fuel consumption and power requirements, monitoring the impact of factors such as hull roughness can be more difficult. While acknowledging this, Sari outlines an approach to assess the impact of changes in hull surface conditions.

Her analysis also benefits from StormGeo’s access to data from about 4,000 vessels, she says in the podcast, but emphasises that it is anonymised so clients need not be concerned that their specific data could be identified. In fact, since the industry is looking for a greener future with lower emissions and more efficient ships, “keeping this knowledge from others would not benefit anyone,” Sari maintains.

Human expertise remains important

Asked about the role of AI in analysing vessel performance data, Sari highlights both its potential and its limitations. AI can help identify patterns and accelerate analysis, but errors in the underlying data and errors generated by AI itself still need to be recognised and challenged. Human expertise therefore remains an important part of the process, with StormGeo also exploring how AI can be applied within its vessel performance tools while keeping data quality and expert oversight at the core.

Alongside her insights into ship performance data, she also comments on some performance-improving technologies, including wind-assistance, air lubrication systems, a novel antifouling technology and future fuels.

All these options provide meaningful savings but these vary between ship types and hull shapes. And they each present specific challenges, so careful assessments should be made to guide investment decisions. “We need to monitor the pioneers of these technologies and operational data will show us how successful each of these technologies will be in the future,” she says.

As the industry moves towards more connected and increasingly automated decision-making, Sari argues that the real value of vessel performance data lies not simply in collecting more of it, but in understanding what it means, knowing what can be trusted and turning that insight into better decisions.