Technology rarely fails on its own. It fails when it lands in an office or on a bridge where nobody was asked what they needed.

The best rollouts begin with a conversation rather than a purchase. Someone describes the daily friction and the tool is shaped around it.

Shipowners and managers have seen every system that was installed and then ignored. The ones worth backing are built with the people who use them.

Columbia Group is using AI for document processing, compliance support and a conversational safety management system. Its approach starts with the business problem rather than the technology, and with staff involved from the outset.

Many shipping companies have been testing AI tools, yet few can point to systems their staff use every day. Columbia Group says it has passed that stage.

Christina Orfanidou, who leads digital and technology initiatives at Columbia Group, said the company has changed direction. “We have now shifted from experimenting that we’ve been doing for a long time to actual deployment of systems,” Orfanidou told Digital Ship.

Early deployment has focused on shore-based operations, covering document processing, compliance support, knowledge management and reporting. The company’s safety management system is AI-assisted, so employees can retrieve corporate and safety procedures through a conversational assistant using natural language.

Her experience of live use is that projects should start with the business problem. “We should not be talking about, ‘I need to put AI there.’ We should really understand the process and the business problem,” she said. Columbia has seen projects underperform when the technology came before an understanding of how people worked, so staff now help define the problem they are meant to solve.

Data is the next hurdle. Columbia holds many structured sources in a data warehouse, but information still sits across multiple systems. “While we are trying to build AI solutions, many of our data issues are surfacing,” Orfanidou said. Combining operational and financial data could help teams trace operational problems back to their causes.

Columbia is also preparing a “citizen AI empowerment” programme, giving employees tools and training to build their own assistants without coding skills.

Onboard, the focus is cautious. “I see a lot of opportunity in decision support, not in autonomous decision making,” she said. Full autonomy remains distant, and she attributes that to regulation rather than technology. “It’s not an issue of technology, it’s an issue of aligning this technology with the regulatory requirements of the shipping sector,” Orfanidou said.

The industry problem here is a familiar one. AI interest is high, but poor data, scattered systems, unclear use cases and staff reluctance keep many projects from reaching daily use. Shipowners and managers weighing AI spend can take from Columbia that the sequence matters: define the problem, involve the people, then fix the data.

Orfanidou will join the AI and automation panel at the Digital Ship Summit in Athens on 15 October. “What I’m excited about is hearing practical deployment stories,” she said.

Data requirements rarely fall evenly across a fleet. One vessel may use significantly more connectivity than another, while requirements can also change over time.

AST Networks offers pooled LEO data options that allow operators to manage connectivity across multiple vessels rather than treating every allowance in isolation.

For shipowners and operators, that creates more flexibility to match available data with actual fleet requirements and adjust as those requirements evolve.

It is part of AST Networks’ wider approach to connectivity: providing operators with choice rather than prescribing a single package for every vessel.

» See how AST Networks approaches flexible fleet connectivity «

Why shipowners must fix data quality before scaling AI

Clarissa Wright examines why reliable, standardised fleet data is essential for effective AI and analytics.

CMB.TECH sends 210,000 dwt ammonia-powered bulk carrier to sea

CMB.TECH has sent Mineral Europa on its maiden voyage to Australia as it plans to add 21 ammonia-powered vessels to its fleet by 2029.

Robosys tests AI navigation against 55 DNV scenarios

Robosys has tested its VOYAGER AI against 55 maritime traffic scenarios at MARIN, examining collision avoidance and compliance with international navigation rules.

MineHub brings real-time vessel tracking to smaller commodity traders

MineHub has added real-time bulk vessel tracking to Jules AI, giving smaller commodity traders visibility of vessel locations, voyage status and estimated arrivals.

Battery-powered fleet surpasses 2,000 vessels

More than 2,000 commercial vessels delivered with batteries or on order represent 2.4 GWh of capacity, according to the Maritime Battery Forum.

Rystad Energy and OceanScore add market prices to FuelEU analysis

Rystad Energy and OceanScore are integrating FuelEU pooling prices into clean shipping research, helping shipowners compare compliance costs and alternative fuel options.

» Secure your Delegate Pass «

» Register your Interest «

InterMaritime Ship Management’s Zenon Oratis explains why shipowners must look beyond cyber compliance to protect connected vessels, maintain operations during attacks and strengthen onboard resilience.