FAQ: Vessel Performance Model (VPM)

What is a Vessel Performance Model (VPM)?

A Vessel Performance Model (VPM) is a computational system that predicts how a specific ship will behave under specific weather and operational conditions. It is also commonly called a digital twin — a virtual replica of the vessel that replicates its real-world performance in software.

Given a set of environmental inputs — wave height and direction, wind, current, sea state — and an engine setting (RPM), a VPM answers:

• How fast will the vessel travel?
• How much fuel will it consume?
• What are the safety and comfort implications of this route?

Every shipping organization has some version of a VPM or digital twin. What differs is how those models are built, what data they rely on, and how accurately they perform outside of calm conditions.

What is a digital twin in maritime shipping?

In maritime shipping, a digital twin is a vessel-specific computational model — also called a Vessel Performance Model or VPM — that replicates how a real ship performs across a range of sea states and engine settings.

The term "digital twin" reflects the goal: to create a living virtual copy of each vessel that mirrors its actual behavior at sea, continuously updated as conditions and vessel characteristics change.

A maritime digital twin is built from:

• Vessel particulars — hull geometry, propeller specifications, engine data, auxiliary propulsion systems such as Flettner rotors
• Historical operational data — how the vessel has actually performed at sea across different conditions
• Real-time environmental inputs — waves, wind, current, sea surface temperature

Unlike generic speed-fuel curves that apply average assumptions across vessel classes, a digital twin is specific to each vessel and updates continuously as the vessel's condition and behavior change over time.

Digital twins power voyage optimization, voyage simulation, charter party tracking, and fleet performance monitoring.

Why do Vessel Performance Models and digital twins matter for maritime shipping?

Vessel performance is never a single number. The same engine setting can produce very different outcomes depending on wind, waves, and current — and those differences compound quickly into fuel burn, arrival time, emissions, and safety risk.

Without an accurate VPM or digital twin, operators are forced to rely on experience, industry averages, or models that work in calm water but degrade as conditions worsen. That gap between assumptions and reality is where costs accumulate and confidence erodes.

Accurate vessel performance modeling and digital twin technology enable:

• Better voyage planning and route selection across all vessel types — dry bulk, tanker, LPG, container, and more
• More reliable bunker cost estimates and VLSFO consumption planning
• Smarter speed optimization and slow steaming decisions to meet charter party terms and laycan windows
• Reduced emissions per voyage, improved CII ratings, and EU ETS cost management
• Defensible data for charter party disputes and off-hire claims
• Improved safety in deteriorating sea states

Why are traditional speed-fuel curves insufficient?

Speed-fuel curves are a widely used industry tool that relate a vessel's speed to its fuel consumption at a fixed Beaufort scale. They are a rough starting point, but they have significant limitations compared to a modern VPM or digital twin:

• They treat headwinds and tailwinds the same, even though they produce opposite effects on vessel performance
• They typically ignore currents entirely
• They represent averages across conditions, not point-in-time predictions
• They can be wrong by 10–15% or more, making it impossible to reliably distinguish between routes that will behave differently at sea

For voyage optimization — which requires comparing thousands of possible routes, many of which differ only slightly — that level of error is too large to identify meaningful savings. A digital twin replaces these static curves with a dynamic, physics-based model that calculates performance at every point along a route, under the exact conditions forecast.

What data does Sofar's Vessel Performance Model use?

Most VPMs are initialized at the start of a voyage using only vessel particulars — static specifications like hull dimensions and engine data. Sofar's approach goes significantly further, and the difference matters for accuracy.

Sofar's VPM draws on three categories of data:

Historical vessel data (used at the start of a voyage — a key Sofar differentiator)

Before a voyage begins, Sofar initializes each vessel's digital twin using actual historical operational records from that specific ship — including past noon reports, historical speed and fuel logs, and observed performance across prior voyages. This means the model arrives at voyage start already calibrated to how that vessel truly behaves, not just how it is specified on paper. Most competitors rely on vessel particulars alone at this stage; Sofar's use of real historical performance data produces a materially more accurate starting point, especially for vessels whose real-world behavior diverges from their specifications due to age, maintenance history, or design variations.

Real-time weather data (used continuously throughout a voyage — another differentiator )

Live ocean and atmospheric observations — wave height, wave period and direction, wind, current, and barometric pressure — are fed into the model continuously as the voyage progresses. This data comes from Sofar's own global Spotter sensor network, producing wave forecasts specifically optimized for maritime routing and tuned to perform well in the heavy sea states where accuracy matters most. Errors in real-time wave data translate directly into errors in predicted speed and fuel burn, which is why the quality of Sofar's in-house forecast is a direct input to VPM accuracy.

Noon reports (used for ongoing daily calibration)

Noon reports submitted by vessel crews provide daily performance data that recalibrates the model continuously throughout the voyage. Because vessel condition changes over time — biofouling, hull cleaning, minor damage — daily recalibration keeps the digital twin aligned with the vessel's current state. Sofar applies data sanitization to filter out manual reporting errors before they can corrupt the model.

This three-layer approach — historical initialization, real-time weather, and daily recalibration — is what allows Sofar's VPM to remain accurate from the first waypoint to the last, across all sea states.

What is a balance-of-forces approach to vessel performance modeling?

A balance-of-forces approach is the physics foundation of a high-quality digital twin. It explicitly models the physical forces acting on a ship under real marine weather conditions, rather than relying on statistical approximations.

Resistive forces (what slows the vessel down):

• Wave resistance
• Air resistance
• Water friction and wave-making resistance
• Current effects

Propulsive forces (what drives the vessel forward):

• Main engine propulsion
• Wind-assisted propulsion (e.g., Flettner rotors, where installed)

For a given RPM, the digital twin solves the physics equation where propulsive and resistive forces balance — producing a precise prediction of speed and fuel burn under those specific conditions, at that specific moment on that specific route.

This approach provides a reliable baseline even when historical data is sparse, and is particularly valuable in rough sea states where pure data-driven methods break down.

How does a hybrid physics and AI approach improve VPMs and digital twins?

A purely physics-based digital twin provides a strong foundation but is subject to bias — real ships vary in ways not fully captured by their reported technical specifications, and those specifications change over time.

A purely data-driven (AI) model is limited by the quality and quantity of available data, and breaks down in sea states underrepresented in historical records.

A hybrid physics and AI approach combines the strengths of both:

• Physics gives a reliable baseline even in conditions rarely observed in historical data (achieving approximately 10% fuel error before calibration)
• AI-driven calibration closes the gap between the idealized model and how a real vessel actually behaves, adjusting for vessel-specific characteristics that cannot be inferred from technical specifications alone
• Daily recalibration keeps the digital twin aligned with the vessel's current condition, automatically accounting for biofouling, hull cleaning, damage, and other changes

This hybrid architecture is why high-quality maritime digital twins significantly outperform both pure physics models and pure data models in real-world conditions.

How does biofouling affect vessel performance — and how do digital twins account for it?

Biofouling is the accumulation of marine organisms (algae, barnacles, mussels) on a vessel's hull over time. It increases hull friction and drag, directly reducing speed and increasing fuel consumption at any given RPM.

Biofouling is one of the most significant causes of performance degradation over a vessel's operational life. Other factors that change performance over time include:

• Hull cleaning and polishing events
• Propeller maintenance
• Minor collision or structural damage
• Changes in draft and cargo configuration

A static digital twin that is not continuously updated will drift away from the vessel's actual behavior as these changes accumulate. A high-quality VPM recalibrates daily against fresh operational data — ensuring predictions remain aligned with the vessel's current state. This is what makes a digital twin "live" rather than a snapshot.

Why does marine weather forecast accuracy affect digital twin and VPM accuracy?

Waves are a primary driver of ship resistance. Small errors in wave height, period, or direction translate directly into large errors in the speed and fuel burn predicted by a VPM or digital twin.

If the digital twin cannot accurately represent the marine weather conditions a vessel is experiencing, it cannot accurately tune the vessel's response to those conditions — and calibration against historical performance becomes unreliable.

This is why marine weather forecast quality is one of the most important determinants of VPM accuracy. A wave forecast that accurately captures:

• Swell (long-period waves generated by distant storms) vs. seas (short-period waves generated by local wind)
• 6-meter vs. 7-meter waves — which have fundamentally different operational implications than the difference between 2-meter and 3-meter waves, because operational limits and safety thresholds often live in that upper range

...produces dramatically more accurate digital twin predictions, especially in the heavy sea states where maritime weather routing decisions are most consequential.

Why does digital twin speed matter for voyage optimization?

Most digital twin and VPM systems run on remote servers and return results in minutes. That latency is acceptable for a single point calculation, but it creates a bottleneck for interactive voyage optimization.

Real-time maritime decision-making requires:

• Comparing multiple vessels for the same fixture simultaneously before committing
• Exploring route alternatives as marine weather forecasts evolve throughout a voyage
• Adjusting RPM guidance and speed optimization in response to changing sea states
• Running deviation analysis instantly when conditions shift — so masters can assess laycan risk on the spot
• Understanding the bunker cost and ETA implications of every routing decision in real time

When a digital twin takes minutes to respond, operators stop asking questions. When it responds in milliseconds, they start exploring — which leads to better decisions and measurable bunker savings.

A digital twin fast enough for interactive use also enables deeper collaboration between human routers and automated optimization systems, allowing both to work together in real time rather than sequentially.

What accuracy should a high-quality maritime digital twin achieve?

Digital twin and VPM accuracy is typically measured by two metrics:

• Speed error (mean absolute error in knots) — how closely the predicted speed matches actual observed speed
• Fuel error (mean absolute percent error) — how closely the predicted fuel consumption matches actual consumption

A physics-only baseline typically achieves approximately 10% fuel error before AI calibration. Industry standard systems using speed-fuel curves may achieve ±10–15% accuracy in typical conditions, degrading significantly in rough weather.

High-quality, calibrated digital twins that combine physics modeling with AI-driven calibration and high-accuracy wave forecasts can achieve substantially lower error rates — enabling voyage optimization algorithms to meaningfully distinguish between routes that would otherwise appear identical.

What is the relationship between digital twin accuracy and voyage fuel savings?

The relationship is direct. Voyage optimization is not about evaluating one route — it requires comparing thousands of possible routes and speed profiles, many of which differ only slightly.

If a digital twin is only accurate to approximately 10%, many routes appear interchangeable. The system cannot reliably identify which routes will actually save fuel, reduce emissions, or improve on-time arrival.

As digital twin accuracy improves:

• Subtle differences in route geometry, timing, and RPM become meaningful
• Optimization algorithms can make sharper tradeoffs between time, fuel, safety, and comfort
• Fuel savings compound across a fleet and over a full year of voyages

This is why vessel performance modeling accuracy is the limiting factor in voyage optimization outcomes — not the sophistication of the routing algorithm.

How are digital twins and VPMs used in voyage optimization?

Digital twins are the computational backbone of voyage optimization. They enable optimization systems to:

• Cost each possible route — by calculating the fuel burn and transit time for every candidate path through the ocean, at every point along the way
• Compare millions of route options — identifying which combination of waypoints, speeds, and timing produces the best outcome given weather, vessel performance, charter terms, and market conditions
• Provide daily RPM and routing guidance — as weather forecasts update and conditions evolve throughout a voyage
• Model deviations in real time — allowing operators to instantly understand whether a course change is permissible and what it will cost

Digital twins also power pre-voyage planning tools (voyage simulators), charter party tracking, speed-fuel curve generation, and fleet performance monitoring.

What is a voyage simulator and how does it use digital twin technology?

A voyage simulator is a planning tool that uses a vessel's digital twin to estimate how that specific ship will perform on a specific route under a specific forecast — before the voyage begins.

It allows commercial teams, operators, and chartering departments to:

• Generate precise bunker consumption and ETA estimates for prospective fixtures — before the vessel is committed
• Compare multiple vessels for the same voyage, accounting for their individual performance characteristics
• Evaluate the impact of different speed profiles, slow steaming options, and routing choices on cost and CII
• Respond quickly to market opportunities and laycan negotiations with confidence
• Build defensible pre-voyage baselines for charter party performance tracking

Because a voyage simulator relies entirely on digital twin accuracy to generate its estimates — including the quality of the marine weather forecast powering it — the fidelity of the underlying vessel performance model determines the reliability of every output.

How does a Vessel Performance Model support CII compliance and emissions reporting?

The Carbon Intensity Indicator (CII) is an IMO rating system that measures a vessel's annual carbon emissions relative to its transport work. As of 2023, vessels must achieve minimum CII ratings — and those thresholds tighten each year through 2030.

A Vessel Performance Model is central to CII compliance in two ways:

Optimizing CII performance during a voyage

Because CII is calculated from actual fuel consumption over the full year, every voyage matters. A high-quality VPM allows operators to:

• Model the CII impact of different speed profiles, including slow steaming, before committing to a route
• Identify bunker-saving routing options that reduce CO₂ output without compromising ETA
• Make daily RPM adjustments that accumulate into meaningful annual CII improvement

Accurate CII tracking and reporting

CII ratings are calculated from reported fuel consumption data. A VPM that tracks predicted versus actual performance throughout each voyage produces the structured data needed for accurate IMO DCS reporting, EU ETS compliance, and end-of-year CII ratings — and provides an auditable record if ratings are disputed.

For fleets managing a mix of vessel types — dry bulk, tanker, LPG, container — a VPM that is vessel-specific rather than class-averaged is especially important, because CII performance varies significantly between individual ships even within the same class.

How does vessel performance modeling help with charter party disputes and off-hire claims?

Charter party disputes — disagreements between shipowners and charterers over vessel performance, speed, and fuel consumption — are one of the most common sources of financial and legal risk in commercial shipping.

Performance claims typically arise when:

• A vessel arrives later than the agreed ETA, triggering demurrage or off-hire
• A vessel consumes more fuel than the charter party warrants
• Adverse marine weather is cited as the cause of underperformance, but the charterer disputes the severity

A vessel digital twin addresses all three scenarios:

Building a defensible performance record

Because a high-quality VPM predicts vessel performance at every point along a route — accounting for actual marine weather conditions experienced — it creates a continuous, data-backed record of how the vessel should have performed relative to the conditions it encountered. This is far more defensible than noon reports alone.

Weather defense documentation

When adverse marine weather is the cause of speed loss or excess fuel consumption, the VPM can demonstrate precisely how much resistance was added by specific wave heights, directions, and sea states — and what the vessel's predicted performance would have been under the charter party's weather warranty conditions.

Separating weather from vessel underperformance

One of the hardest problems in charter party disputes is separating genuine vessel underperformance (e.g., due to biofouling or mechanical issues) from weather-induced speed loss. A digital twin that models both factors independently — updated daily with the vessel's actual performance data — makes this distinction quantifiable and auditable.

How does Sofar Ocean's Wayfinder use digital twin and VPM technology?

Sofar Ocean's Wayfinder platform is built on a vessel performance modeling and digital twin system designed to be accurate across all sea states and fast enough for real-time, interactive use.

Sofar's approach combines:

• Physics-based digital twin modeling — a balance-of-forces framework that provides a reliable baseline even when operational data is limited
• AI-driven daily calibration — each vessel's digital twin is automatically recalibrated every day as fresh operational data arrives, keeping predictions aligned with the vessel's current behavior
• High-accuracy marine weather forecasts — powered by real-time observations from Sofar's global Spotter buoy network, producing maritime weather forecasts specifically optimized for vessel routing and tuned to perform well in the heavy sea states where accuracy matters most
• In-browser execution — VPM and digital twin calculations run directly in the user's browser in approximately 20 milliseconds, enabling real-time exploration of route alternatives without server round trips

This combination enables voyage optimization, voyage simulation, charter party tracking, and fleet performance monitoring at the accuracy and speed that modern maritime operations require.

‍