The levels of fidelity
"Digital twin" is used to describe everything from a 3D model of a ship to a live physics-based simulation fed by sensor data. The gap between those two is enormous, and most disappointment with digital twins comes from buying the first while expecting the second.
| Level | What it is | Fed by | Useful for |
|---|---|---|---|
| 1. Digital record | Structured data about the ship: equipment register, drawings, certificates | Manual updates | Documentation, handover, class |
| 2. Digital model | Geometry and system models — 3D, P&IDs, structural models | Design data | Modification planning, retrofits |
| 3. Digital shadow | Model updated with operational data | Sensors, one-way | Condition tracking, performance |
| 4. Digital twin | Model that simulates behaviour and is continuously synchronised | Sensors, two-way | Scenario testing, prediction |
| 5. Predictive twin | Twin plus forecasting of future states | Sensors plus models | Remaining useful life, planning |
The bars represent relative cost; the line represents realised value across a typical commercial fleet today. Note that the value line peaks in the middle. For most operators, a well-maintained digital record plus a digital shadow delivers more than an ambitious twin that nobody has the data quality to feed.
What a level-1 digital record actually delivers
This is the unglamorous one, and it is where most fleets should start:
- Complete equipment register with hierarchy, shared by PMS, procurement and inventory
- Drawings, manuals and system diagrams, current and retrievable
- Certificate register with structured expiry data
- Maintenance and repair history, including dry dock records
- Modification history — what has actually changed since delivery
- Inventory of Hazardous Materials, kept current as required under the Hong Kong Convention regime
Every one of those has an immediate operational use, and together they are the foundation any higher level depends on. A fleet that cannot produce a correct equipment register cannot build a twin.
Where higher fidelity genuinely pays
Hull structural monitoring on large vessels, where strain gauges and structural models together inform loading decisions and fatigue assessment.
Complex machinery systems — LNG cargo systems, DP systems, propulsion trains — where simulating a scenario before doing it has real safety value.
Retrofit planning. Modelling an energy saving device, a shore power connection or a fuel conversion before committing capital.
Newbuilding handover. A yard delivering a structured digital record with the ship, rather than boxes of PDFs, saves years of reconstruction.
Training. Simulation of vessel-specific systems for crew familiarisation, particularly on alternative-fuel tonnage where the competency gap is widest.
Where it does not pay yet
- General cargo and dry bulk tonnage with limited instrumentation and thin margins
- Fleets whose master data is inconsistent between vessels
- Any application where the decision it would inform is already made adequately by an experienced engineer
- Anywhere the twin would be built once and never updated — a twin that diverges from the ship is worse than no twin
The maintenance problem
A digital twin is only true on the day it is synchronised. Ships change: equipment is replaced with non-identical items, piping is modified, sensors fail and are bypassed. Keeping a twin accurate requires a process — change control that updates the model whenever the ship changes — and that process costs money indefinitely.
Before buying, ask: who updates this after every dry docking, and how is that funded? If there is no answer, you are buying a snapshot.
A pragmatic approach
| Step | Action |
|---|---|
| 1 | Fix the equipment register and certificate data across the fleet |
| 2 | Capture drawings and modification history in a retrievable structure |
| 3 | Add operational data from existing sensors — the digital shadow |
| 4 | Model one high-value system where simulation changes a decision |
| 5 | Establish change control that keeps the model synchronised |
| 6 | Only then consider full twin fidelity, and only where the value is demonstrable |
vendor-neutral guidance; IHM obligations per the Hong Kong Convention regime in force since 26 June 2025. Cost/value chart is conceptual. Reviewed by the Zeaclub Editorial Team, 24 August 2026.
Frequently asked questions
What is a digital twin of a ship?
A digital representation of the vessel that is synchronised with its real-world state through operational data, and which can be used to analyse or simulate behaviour. Simpler digital records and models are often described by the same term.
Do digital twins reduce maintenance cost?
They can, where they support condition-based maintenance on high-consequence machinery. The saving comes from the decisions, not from the model itself.
Do I need one?
Most operators get more value from a complete, accurate digital record and a data-fed digital shadow than from a full simulation twin.
What is the biggest risk?
Divergence. A model that is not maintained after modifications and dockings becomes misleading, which is worse than having no model at all.