Marine

AI costs and ROI for marine

AI costs are not the licence fee. They are the people, the data, the integration, the governance, and the ongoing maintenance. A marine board that only sees the headline price is seeing half the picture.

By Zakir Hoosen, Director, Fuzzelogic Solutions. Board-level guidance in plain English.

Every AI pitch I see in the marine sector starts with the price tag. The platform costs this much. The implementation costs this much. The savings are this much. The numbers look good on a slide.

Then six months in, the real costs appear. The data was not ready, so there was a data project nobody budgeted for. The system did not integrate with the existing maintenance platform, so there was an integration project nobody budgeted for. The crew needed training, so there was a training programme nobody budgeted for. The governance framework did not exist, so there was a governance project nobody budgeted for.

This article is about what AI actually costs a marine company, how to think about ROI honestly, and where the money usually goes wrong.

The headline cost is not the real cost

When a vendor quotes a price for an AI system, that price covers the software and usually a basic implementation. It does not cover the things that make the difference between success and failure.

The real cost of AI in a marine company has five components.

1. The platform licence or subscription fee.
2. The data preparation and integration work.
3. The people, internal and external, to build and maintain it.
4. The governance and compliance framework.
5. The ongoing maintenance, updates, and monitoring.

In my experience, the platform licence is roughly a quarter of the total cost over three years. The rest is the work around it. A marine company that budgets only for the licence is budgeting for a quarter of the project.

Why marine costs are higher than office-based costs

A shipping company is not an office. The data is scattered across vessels, shore offices, and third-party systems. Connectivity is limited at sea. Crew change regularly. The systems on board were not designed to integrate with AI platforms.

These factors add cost. Data extraction from legacy vessel systems takes longer. Integration with operational technology on board requires specialist knowledge. Training crew who rotate every few months requires a different approach to training office staff. Governance must account for systems operating in remote locations with limited oversight.

The Isle of Man Ship Registry does not prescribe which systems vessel operators use. That flexibility is commercially valuable. It also means there is no standard data format across the fleet. Each company must map its own data landscape, and that takes time and money.

ROI in the marine context

Return on investment is not a single number. It is a story about what changes and what it costs to make that change happen.

The honest ROI calculation for a marine AI project should include the cost of the data work, the cost of the people, the cost of the governance, and the cost of maintenance over at least three years. It should also include the cost of failure. What happens if the system is wrong? What does a wrong decision cost in cargo, in fuel, in crew time, in reputation?

The two thirds that did not deliver did not fail because the technology was wrong. They failed because the ROI was assumed, not calculated, and the costs were underestimated from the start.

The 10-20-70 problem

Seventy percent of the value is in people and processes. That means seventy percent of the cost is also in people and processes. A marine company that only budgets for technology is missing the largest part of the picture.

In practical terms, this means the training budget should be at least as large as the technology budget. The change management budget should be larger. The governance budget should be built into the project from the start, not added as an afterthought.

What a board should ask about cost

When the AI proposal lands on the board table, ask these questions.

First, what is the total cost over three years, not just the first year? AI is not a one-time purchase. It is an ongoing commitment.

Second, what is the cost if it fails? Not the technology failing, but the project failing. What have you spent and what do you have to show for it?

Third, who has calculated the data cost? If nobody has, the data cost has not been estimated, and it is always larger than expected.

Fourth, what is the training budget? If the system is used by crew, by shore staff, or by commercial teams, someone has to train them. That takes time and money.

Rushing the cost discussion is how projects go over budget. Taking an extra week to get the numbers right saves months of explaining overruns.

The Isle of Man context

The Island has a marine sector that punches above its weight. Ship managers,注册 agents, law firms, and classification societies all operate here. The commercial value of the Island's maritime reputation is real. AI that is implemented well and governed well adds to that reputation. AI that is implemented poorly and governed not at all damages it.

Boards on the Island should think about AI cost not just in terms of their own return, but in terms of what it means for the registry and for the Island's standing in the global marine market. That is not charity. It is commercial self-interest.

The honest version

Fuzzelogic is an Isle of Man firm that has spent nineteen years modernising platforms across banking, insurance, healthcare, retail, manufacturing, and government. We tell boards what most consultants will not: the honest cost of AI is always higher than the headline, and the honest return is always lower than the pitch. We would rather give you the real numbers and let you make an informed decision.

Start with the assessment. Two to four weeks, fixed price, and you own the verdict and the roadmap whether or not we build any of it. When you are ready to talk AI, call Fuzzelogic Solutions and ask for Zak. www.FuzzelogicSolutions.com | info@FuzzelogicSolutions.com | +44 (0)1624 618950