From Data to Decisions: METIS Seeks the Intelligence Behind the Numbers

For all of the talk about digitalization in shipping, one fundamental truth is fundamental: collecting data is no longer the hard part.

Ships today are awash in information. Engines, sensors, navigation systems and other onboard equipment continuously generate streams of operational data, creating an enormous digital record of how a vessel performs. The challenge is figuring out what all of that information actually means — and, more importantly, what a shipowner or operator should do about it.

That is the space in which METIS has built its business.

Founded in 2016 and headquartered in Athens, Greece, METIS is part of the ERMA TECH GROUP and has developed a fleet performance management platform around a telemetry-first philosophy. Today, the company has more than 500 contracted vessels worldwide, processing more than 12 billion data points every month. Its historical data foundation is approaching one trillion records.

For Ioannis Stratakos, who earlier this year assumed the role of CEO after serving as the company's chief technology officer, the sheer volume of data is less important than what can be extracted from it.

“Technology only creates value when it solves real business problems,” Stratakos said. That philosophy increasingly defines METIS's approach: move beyond dashboards and toward intelligent systems that can help operators make confident, timely and economically sound decisions.

Image courtesy METIS

Dashboards answer questions while intelligent systems tell you which questions to ask.”

- Ioannis Stratakos,
CEO, METIS

Beyond the Dashboard

Stratakos brings an engineering and technology background to the challenge. An electrical computer engineer educated at the National Technical University of Athens and the Technical University of Munich, he joined METIS more than eight years ago, when digitalization in shipping was still gathering momentum.

During his tenure, he has worked across product development, cloud architecture, customer engagement and strategic partnerships, eventually leading the company's technology evolution and AI strategy.

That evolution mirrors the broader transformation taking place in maritime.

“Most vessels are highly connected, continuously generate vast amounts of information from onboard equipment and sensors,” Stratakos said. “But the challenge is no longer gaining access to the data. It's understanding what the data actually means and more importantly, what action to take as an operator.”

A sensor reading or KPI viewed in isolation provides only part of the story. Proper interpretation requires context: weather, loading condition, vessel configuration, machinery status and even the commercial objectives of a particular voyage.

It also requires semantics—understanding what each measurement represents and how it relates to data coming from other systems.

Then there is fragmentation. Information and expertise can be scattered among engineering, performance, compliance and commercial departments. No single individual can realistically synthesize all of those inputs continuously.

METIS sees AI as an opportunity to bridge those gaps.

Rather than simply analyzing individual datasets, an AI-enabled system can correlate information across domains, compare current vessel behavior with historical experience and identify relationships that might otherwise remain hidden. The result, Stratakos argues, is an operational reasoning engine that augments the expertise of fleet teams.

The Trust Factor

Before any of that can happen, however, operators have to trust the information.

That may sound obvious, but it is one of the central challenges of turning maritime data into something useful. Sensors drift. Communications fail. Equipment goes offline. Data gaps occur. Yet operational decisions cannot wait for a theoretically perfect dataset.

“The challenge isn't achieving perfect data,” Stratakos said. “It's understanding the quality of the information available and how much confidence you have in it.”

METIS therefore puts significant emphasis on resilient IoT architecture, continuous data validation and what it calls its “data health framework.” The system evaluates the completeness, validity and reliability of individual data streams and quantifies the confidence associated with them.

That confidence becomes part of the decision-making process.

If a recommendation is based on highly reliable data, an operator can act decisively. If confidence is low, the operator may choose to validate the situation before acting.

This distinction is important. The objective is not simply to deliver another KPI. It is to provide context around the KPI and establish whether it is sufficiently trustworthy to support an operational decision.

Stratakos summarizes the philosophy succinctly: “Dashboards answer questions while intelligent systems tell you which questions to ask.”

Turning Intelligence Into Dollars

For shipowners, digitalization ultimately has to connect to the bottom line. METIS identifies three areas where its technology can directly influence profitability: maintenance planning, commercial performance and voyage execution.

Consider hull and propeller fouling. Cleaning too early can waste money, while cleaning too late can result in unnecessary fuel consumption. By continuously monitoring vessel performance, METIS seeks to identify the optimal intervention point and quantify the financial consequences of fouling.

Commercial performance presents another opportunity. By establishing a more accurate understanding of how an individual vessel performs under different operating conditions, owners can develop more realistic charter party terms. That can reduce the risk of performance claims while preventing owners from becoming unnecessarily conservative and potentially less competitive.

And when a claim does arise, measured operational data can help separate genuine underperformance from factors such as weather or operational constraints. Stratakos points out that a single performance claim can reach hundreds of thousands of dollars or more, meaning that avoiding one unjustified claim, or negotiating a better charter arrangement, can potentially deliver an immediate return on the digital investment.

Voyage execution is another major focus. Rather than treating a voyage plan as static, METIS can continuously evaluate weather forecasts and required arrival times and recommend adjustments to speed and routing.

The goal is straightforward: arrive safely and on time while consuming the least amount of fuel consistent with the commercial commitment.

Every Ship Has a Fingerprint

One of the more compelling aspects of METIS's approach is its recognition that ships are not interchangeable, even when they are sister ships.

Over time, vessels develop what Stratakos calls their own “fingerprint.” Maintenance history, coating condition, fouling exposure, machinery wear, trading patterns and operating practices all influence how an individual vessel behaves.

METIS addresses that challenge through a “light gray” modeling approach, combining physics-based principles with continuous calibration from real operational data.

The models begin with vessel-specific information, including sea trials, and are continually refined using measurements from onboard systems. Naval architecture and engineering principles provide the foundation, incorporating factors such as resistance, propeller characteristics and weather effects, while operational data allows the model to reflect how the vessel actually behaves in service.

The philosophy is that a vessel should be measured against its own expected behavior, rather than simply compared with a generic peer.

As Stratakos put it, sister ships may share a hull form and machinery, “but they do not share a life.”

The Digitalization Journey Is Really About People

Perhaps the most significant point in Stratakos's vision is that AI is not intended to replace maritime expertise. It is intended to make that expertise more scalable.

METIS had envisioned a “cyber assistant” for maritime operations long before generative AI became mainstream. The emerging generation of AI tools now makes it increasingly practical to create digital teammates capable of understanding operational context, reasoning across domains, explaining recommendations and proactively engaging users.

That has particular significance at a time when shipping is wrestling with an aging workforce and a continuing need to attract, develop and retain experienced personnel.

The knowledge accumulated by an experienced engineer, superintendent or fleet manager has traditionally resided largely in that individual's head. AI creates the possibility of capturing, refining and making portions of that expertise available across an entire fleet.

But technology, Stratakos cautions, is only part of the equation.

Shipowners should not simply buy more software. They should build a trusted digital ecosystem in which navigation, weather, engine, performance and other specialized systems can work together, with an intelligence layer connecting the information.

And organizations must change along with the technology.

“People” ultimately drive transformation, Stratakos said, noting that successful data-driven decision-making requires cultural change, trust in data and a willingness to challenge established ways of working.

That may be the bigger opportunity—and challenge—for maritime digitalization.

The industry does not need more information for information's sake. It needs better decisions.

For METIS, the path from trillions of data records to measurable business value runs through trusted data, vessel-specific intelligence, AI-enabled reasoning and, ultimately, the people who decide what to do next.

Digitalization, Stratakos said, “is not a technology journey. It's a business transformation journey.”

For an industry increasingly measured in fuel consumed, emissions generated, downtime avoided and dollars earned or lost, that distinction could prove to be the most important data point of all.