LionRock Maritime’s Fuel Measurement Approaches

Method 1: Clamp‑on Flow Meters (External Fuel Meters)

Overview:
Clamp‑on flow meters are easily mounted on existing fuel lines and measure fuel flow without cutting into the vessel’s fuel infrastructure. These are typically used where operators want real‑time and accurate fuel consumption without permanent installation.

Technical Considerations:

• Non‑intrusive installation
• Real‑time fuel flow readings
• Works with legacy vessels
• Often provided/served via LionRock + Techbinder telemetry

Benefits:

• High fidelity measurement
• No downtime for engine disassembly
• Good choice for pilot projects or selective instrumented tugs

Challenges:

• May still require calibration
• Can be sensitive to vibration, line size, and flow regime

Supported in:
FuelExplorer / Tugboat Fuel Index via telemetry ingestion

Method 2: In‑Line Flow Meters (Coriolis, Mass Flow, etc.)

Overview:
In‑line meters measure fuel flow directly within the fuel line. These include Coriolis or mass flow meters that provide volumetric or mass flow data with high precision.

Technical Considerations:

• Requires fuel system modification
• Often best installed during maintenance periods

Benefits:

• Highest possible accuracy
• Stable long‑term measurement
• Well suited for high accurate compliance and auditing

Challenges:

• Downtime needed for installation
• Higher upfront cost

Supported in:
FuelExplorer / Tugboat Fuel Index through Techbinder hardware integration

Method 3: Engine‑Derived Estimation (RPM / Load)

Overview:
Some engines provide internal fuel‑use metrics based on RPM, load, and OEM fuel‑consumption tables (SFOC curves). This is an indirect method but leverages existing engine data.

Technical Considerations:

• Works off engine control / telematics data
• Accuracy improves with consistent load monitoring

Benefits:

• No hardware meter needed
• Uses built‑in engine signals
• Works for engines that support fuel usage outputs

Challenges:

• Not all engines output accurate fuel data, especially older units
• OEM SFOC tables are designed for ideal conditions, not complex tug operations
• Engine load does not equate perfectly to fuel flow without calibration

Supported in:
FuelExplorer / Tugboat Fuel Index when engine RPM / load telemetry is available

Method 4: LionRock ML Model - AIS + Engine Data

Overview:
LionRock’s machine learning model ingests AIS dynamic data (speed, course, route) combined with engine data (RPM or load) to estimate fuel consumption.

Key Inputs:

• Speed over ground / movement pattern
• Engine RPM / load
• Tug class / hull / engine type
• Sail mode (idle, transit, tow‑on)

Benefits:

• Near‑meter accuracy (industry trials showed ~1.8% average deviation)
• Low‑impact deployment (no hardware meters required if engine data available)
• Handles complex operational context

Challenges:

• Needs consistent engine RPM or load data feeds
• Still an estimation model (relies on data quality)

LionRock + Techbinder:
LionRock Maritime’s partnership with Techbinder enables real‑time ingestion of engine data via IoT data loggers, enabling this model to deliver high‑precision estimates without expensive flow meter installations.

Method 5: LionRock ML Model - AIS‑Only Contextual Estimation

Overview:
When engine telemetry is absent, LionRock can estimate fuel consumption purely from operational context analysing AIS and contextual data(speed profiles, job specifics and environmental conditions). The contextual model uses advanced patterns and correlations derived from large datasets to estimate fuel use.

Key Inputs:

• AIS speed & route segments
• Time under tow vs idle vs transit
• Operational state from job logs (via TugIO)

Benefits:

• No hardware or engine telematics necessary
• Deploys fleet‑wide instantly
• Generates normalized metrics (fuel‑per‑NM, fuel intensity)

Challenges:

• Less precise than AIS+engine model, but highly cost‑effective
• Accuracy depends on quality of AIS and job context feeds

Supported in:
FuelExplorer / Tugboat Fuel Index and integrated with PowerCaptain for behavioral context

Proven Results

• LionRock’s contextual fuel estimation models have been demonstrated in industry trials, including Fairplay Towage in Rotterdam, with results showing:

  • Significant insights into fuel usage patterns across idle, transit, and job segments
  • Low cost for getting fuel consumption data of tugboats

  These results validate the combined use of AIS + engine data ML models as a cost‑effective alternative to hardware flow meters in real operations. 

Step 2: Select Measurement Methods

Choose from:

1. Clamp‑on meters
2. In‑line flow meters
3. Engine‑based estimation
4. LionRock Model using contextual and engine data
5. LionRock Model using contextual data only

Where possible, combine methods to improve accuracy and validation (e.g., engine telemetry + AIS).

Step 5: Translate into Insights

Use the outputs to:

• Improve captain fuel awareness
• Correct inefficient behaviors (e.g., excessive speed)
• Support emissions reporting (CII/SEEMP)
• Make better pricing decisions with accurate cost insight
• Drive operational optimization

Evidence & Governance

LionRock’s approach integrates structured measurement with analytical governance:

• Technical partnership: Techbinder IoT for hardware telemetry
• ML model validation: Accuracy tested against real meter data
• Platform normalization: FuelExplorer and Tugboat Fuel Index for comparable metrics
• Behavioral analytics: PowerCaptain for speed and context links to fuel use
• Operational integration: TugIO job state for segment attribution

Key KPI Definitions

• Fuel‑per‑Nautical Mile (NM): Liters consumed per nautical mile for a specific segment
• Contextual Fuel Use: Fuel estimate adjusted for segment (idle / transit / assist)
• Deviation from Meter (%): Model vs actual meter difference
• Normalized Fuel Index: Aggregated measure across fleet, vessel, or captain