For many shipping companies, digitalisation begins with dashboards, AI pilots and connected vessels. For Panagiotis Gkotsis, IT Director at Dorian LPG, the work starts much earlier. Before organisations can expect meaningful results from analytics or artificial intelligence, they need reliable data, clear ownership and a way to measure whether a project has actually delivered value.
Speaking with Digital Ship, Gkotsis drew on almost two decades in maritime IT to discuss the practical obstacles that continue to slow digitalisation, from inconsistent vessel data to unrealistic expectations around AI.
His message? Shipping has work to do in making technology usable.
Data quality remains digitalisation’s biggest obstacle
Dorian LPG has invested in digital projects for many years, using performance analytics, cloud infrastructure and data platforms to support operational decisions. However, Gkotsis said obtaining usable information from ships remains one of the industry’s hardest challenges.
“There are a lot of issues with the quality of data,” he said. “You have a lot of installations where you get different kinds of data for the same exact sensors.”
The problem is not just hardware. Different equipment suppliers, and sometimes different installations from the same supplier, produce inconsistent data structures. That makes fleet-wide analytics far more difficult than expected.
Dorian LPG has attempted several approaches, including middleware and data normalisation tools. While some methods worked, they demanded substantial engineering effort.
“It did work, but it took enormous effort and resources to accomplish it,” he explained.
Replacing every system is rarely practical, particularly for operators with large fleets containing vessels of different ages. Instead, companies often have to build their own normalisation layers before analytics or AI can deliver useful results.
Even modern tools designed to identify faulty sensors or improve data quality still require one essential ingredient.
“In order to do that, you will have to have normalized data initially.”
Gkotsis believes wider adoption of common protocols across maritime equipment would remove much of this complexity. Standardised data formats would make integrations easier, improve interoperability between systems and reduce the effort needed to build digital services across mixed fleets.
Success should be measured before projects begin
While AI continues to dominate industry discussions, Gkotsis believes organisations should first decide what success actually looks like.
“There is an over excitement on what digitalization can achieve,” he said. “You should have criteria, you should have KPIs. You should be able to monitor… What is my success criteria?”
He argued that many AI projects begin with ambitious expectations without defining how improvements will be measured. Vendors often promise lower costs or greater efficiency, yet shipping companies also have a responsibility to define the outcomes they expect before any implementation begins.
“You promised me that AI is going to solve this kind of problem more efficiently. But how do you define it more efficiently?” he said. “This needs to be quantified.”
For Gkotsis, AI remains a decision support tool rather than a replacement for human judgement.
“It is an informed decision,” he said. “Somebody will have to take the ownership of the decision and the liabilities for the decision.”
He also argued that governance deserves more attention than it often receives. Ownership of vessel data remains a contractual issue, with shipowners frequently having to negotiate access to information generated by their own assets.
“You have to fight for the ownership of the data,” he said. “This is ridiculous, but this is the truth.”
Alongside contractual ownership, organisations also need clear internal responsibility for business processes, access rights and the way operational data is managed across departments.
Digital Ship Summit and what actually works
As discussions prepare for the Digital Ship Summit, Gkotsis hopes the industry spends less time discussing future concepts and more time sharing practical experience.
“I would like to hear about the reality,” he said. “I would like to hear about real problems, real solutions, real workarounds.”
He believes honest conversations about projects, including the challenges encountered during implementation, will be more valuable than presentations centred on ideas that have yet to prove themselves in operation.
One topic he is especially interested in discussing with peers is resourcing.
“How do they handle resources,” he said. “All of our IT departments are short on resources.”
Finding and retaining experienced IT and cyber security professionals remains difficult across shipping, particularly for teams expected to support vessels around the clock while managing a wide range of technologies and operational requirements.
For Gkotsis, digitalisation is less about finding the next technology and more about getting the fundamentals right. Better data, measurable objectives and open discussions about real operational experience will do more for the industry’s digital progress than another wave of ambitious promises.
