Shipping’s AI challenge is no longer artificial intelligence. It’s data
- Shipping’s AI problem is no longer the technology itself, but poor data quality and fragmented systems
- Practical AI applications are already boosting productivity, maintenance and administrative efficiency across the industry
- As investment accelerates, industry leaders say success will depend on digitalisation, trust and organisational change rather than better algorithms
AI is already delivering measurable gains across shipbuilding and vessel operations, according to industry leaders gathered for shipping’s annual tech confab. But fragmented data and incomplete digitalisation are preventing wider adoption. As investment pours into maritime AI, the sector’s biggest challenge may be organisational readiness rather than technology itself
FOR many years, AI has been touted as shipping’s next revolution. But as industry leaders gathered in Hamburg for SMM, a more pragmatic conclusion emerged: the technology is largely ready. The industry’s data infrastructure is not.
Across discussions involving shipowners, shipbuilders, class societies and technology providers, executives pointed to growing evidence that AI is already improving maintenance, production, analytics and administrative workflows. Yet fragmented systems, poor data quality and uneven digitalisation remain the biggest obstacles to wider adoption.
The shift in tone was notable. For much of the past decade, maritime AI discussions centred on autonomous vessels and digital disruption. Today, the focus is on applications already generating measurable returns and on the organisational changes needed to deploy them at scale.
As DNV group president and chief executive Knut Orbeck-Nilssen warned during SMM’s opening conference, AI’s success depends on the quality of the data that underpins it.
That message recurred throughout the event. Industry leaders increasingly agree that shipping’s AI challenge is fundamentally a digitalisation challenge.
AI is already helping companies unlock information buried within existing systems, said Oldendorff Carriers managing director of shipbuilding Scott Bergeron. However, fragmented and inconsistent data remains a persistent constraint.
“It’s the old problem: garbage in, garbage out,” he said, noting that companies still spend significant resources cleaning and standardising information before meaningful AI-driven decisions become possible.
ABS Group vice-president of the AI Center Matthew Mower similarly argued that the main obstacles are organisational rather than technological.
He identifies analytics, knowledge management and automation as AI’s three principal value drivers. Yet many companies discover their ambitions exceed the maturity of their underlying data infrastructure.
“What we find, as we have deeper conversations, is they’re really not ready for it, either from a data quality standpoint, or because they lack the technology foundation needed to make data accessible.”
In many cases, the first practical use of AI is improving data quality itself.
“We have to work on the foundation first,” Mower said. “It’s crawl, walk, run stuff.”
HD Hyundai AI strategy lead Aerin Jungmin echoed that assessment, arguing many companies are attempting to accelerate AI adoption before completing their digital transformation programmes. Data remains fragmented, processes vary across sites and critical information is often difficult to access.
Measurable returns are already emerging
While the industry continues to debate AI’s long-term potential, practical applications are already delivering tangible results.
HD Hyundai said nearly 100 AI-powered welding robots are operating at one of its shipyards, boosting productivity by 15%-20%, reducing block production times from 60 days to five days and cutting welding defects by 90%.
Elsewhere, Everllence chief technology officer Gunnar Stiesch highlighted AI’s growing role in condition monitoring and predictive maintenance.
“It’s strongest where an experienced person with domain knowledge comes together with AI algorithms and trusted data,” he said.
Some of the fastest returns are coming from less visible applications. Complexio director Mark Robertshaw said companies are increasingly deploying AI to automate invoice processing, ETA consolidation, laytime calculations and demurrage management.
“The boring bits” often offer the strongest business case, he said.
Investors see shipping as a new AI opportunity
Those real-life cases are attracting fresh capital.
According to TMV co-founder and managing member Marina Hadjipateras, attitudes towards maritime technology have shifted dramatically in recent years.
“All the stakeholders are coming together and embracing it, which is something that five years ago you couldn’t really say,” she said, describing innovation as being “at the forefront of everyone’s minds today”.
The change is drawing expertise and investment from outside traditional maritime circles.
“We’re also seeing founders coming from outside the industry,” Hadjipateras said.
“You’ve got ex-Scale AI and ex-SpaceX people paying much closer attention to maritime, either founding companies or joining early-stage ventures because they see this as a transformational moment for the industry.”
Maritime is increasingly viewed as one of the last big industrial sectors yet to undergo large-scale digital disruption. However, Hadjipateras cautioned that technology alone will not determine success.
“You can’t have just good technology. You need organisational readiness as well. You need the right relationships and partnerships. Trust remains one of the most important factors in shipping.”
Beyond autonomous ships
Perhaps the clearest sign of how the debate has matured is the declining focus on autonomy.
For years, autonomous vessels dominated maritime AI discussions. At SMM, speakers instead highlighted digital twins, predictive maintenance, cybersecurity and decision-support systems.
“We wouldn’t try for full autonomy anyway,” said Meyer Turku chief strategy and transformation officer Anu Ahola.
Jungmin likewise argued that while AI is already improving design productivity, manufacturing efficiency and quality control, autonomous operations remain a much more distant prospect in a safety-critical industry where accountability and trust are paramount.
Mower described AI adoption as a long-term organisational transformation rather than a purely technological challenge, requiring companies to redesign processes while ensuring humans retain responsibility for critical decisions.
From experimentation to operations
The clearest indication that AI is entering a new phase may be the growing expectation that it will soon be embedded in day-to-day operations.
Orca AI co-founder and chief executive Yarden Gross predicted that AI-driven automation will become a routine part of shipping workflows within two years.
Whether that timeline proves accurate remains uncertain. But the message from SMM was remarkably consistent. AI is no longer a future technology waiting for a breakthrough. It is already delivering value across shipbuilding, vessel operations and commercial management.
The industry’s challenge is no longer developing better algorithms. It is building the data quality, system integration and organisational trust needed to put them to work.
