NoonKit

Open, methodology-transparent analysis of ship noon reports — vessel performance and IMO CII compliance, in one toolkit.

Every shipping company collects noon reports daily. The analysis of that data — speed–consumption curves, hull-fouling detection, carbon-intensity compliance — is locked inside expensive proprietary platforms. The open ecosystem has a data standard (the Smart Maritime Network's SVD for Noon Reports) and academic ML models, but no open, auditable analysis tool that sits between them. noonkit fills that gap.

It does two things, both with fully documented methodology:

1. Vessel performance — fits speed–consumption curves, establishes a clean-hull baseline, and quantifies fouling / degradation as a trend over time, with good-weather filtering as standard.
2. IMO CII — computes attained and required Carbon Intensity Indicator and the A–E rating, implemented straight from the MEPC resolutions with a full audit trail and explicit handling of the not-yet-adopted post-2026 factors.

Scope and honesty. Outputs are estimates for planning and insight, not a verified Statement of Compliance. Compliance ratings are issued by Administrations and Recognized Organisations against verified DCS data. noonkit is built to be correct against the published formulae (see the test suite), but it is a decision-support tool, not a regulatory authority.

New to the project or to maritime data? Start with the plain-language FAQ — it assumes no shipping or software background. For formula-by-formula detail see the methodology, and for the audit of every constant against the IMO resolutions — including two transcription errors it caught in circulating secondary sources — see the verification.

• The one existing open CII implementation is single-language, lightly maintained, and CII-only. noonkit combines CII with the performance analysis that operators actually run day to day, in Python, with a real test suite validated against the IMO worked examples.
• It consumes data shaped like the open industry standard (SVD), rather than inventing yet another schema. The ingestion layer maps real-world messy column names onto a canonical schema and surfaces data-quality problems instead of hiding them.
• The methodology is written out formula by formula, each tied to the MEPC resolution it comes from. You can audit every number.

pip install -e . # from a clone
# or, once published:
# pip install noonkit

Requires Python ≥ 3.10, pandas, numpy.

# Generate a demo dataset (with an embedded fouling signal)
python examples/generate_sample_data.py
# Analyze it: performance + CII
noonkit analyze examples/sample_noon_reports.csv \
 --ship-type tanker --dwt 110000 --gt 60000 --year 2024

Example output:

[ Performance ]
 Speed-consumption model: FOC = 0.01524 * speed^2.995 (R^2=0.984, n=17, ...)
 Mean deviation from baseline: +6.81%
 Fouling trend: +1.81 pp / 30 days
[ CII compliance ]
 2024 tanker: rating C (attained 3.831, required 4.103, A/R 0.934)
 Compliant (C or better): True

▶ Try the live dashboard — upload a CSV or click "Use demo data," no install needed.

For a point-and-click version — upload a CSV, see the fouling chart and CII rating in your browser — there's a Streamlit dashboard:

pip install streamlit matplotlib
streamlit run app.py

It opens in your browser, includes a "Use demo data" button, and auto-maps common noon-report column names. The dashboard performs no calculation of its own — every number comes straight from the tested noonkit engine, so the UI can never silently disagree with the library or the CLI.

A hosted version can be deployed free on Streamlit Community Cloud by pointing it at this repo and app.py.

from noonkit import (
 ingest_csv, fit_speed_consumption, baseline_deviation,
 calculate_cii, FuelConsumption, ShipType,
)
# --- performance ---
data = ingest_csv("reports.csv")
excl = data.exclude_mask() # drop implausible rows
model = fit_speed_consumption(data.df, baseline_days=30, exclude_mask=excl)
dev = baseline_deviation(data.df, model, exclude_mask=excl)
print(model.summary(), dev.mean_deviation_pct, dev.trend_pct_per_30d)
# --- CII ---
result = calculate_cii(
 ship_type=ShipType.TANKER,
 dwt=110_000, gt=60_000,
 distance_nm=35_654,
 fuels=[FuelConsumption("HFO", 4_825)],
 year=2024,
)
print(result.summary(), "compliant:", result.is_compliant)

noonkit maps common noon-report column names automatically (e.g. Obs Speed, Distance Obs, ME FOC, AE FOC, Wind Force). The canonical fields follow the spirit of the SVD standard. See examples/sample_noon_reports.csv for a working example and src/noonkit/ingest.py for the full alias table.

pytest

The suite pins the CII engine to the IMO worked examples and the published reference tables, and verifies the performance fit recovers a known power law and a known fouling trend from synthetic data.

Being clear about what this does and doesn't do matters more than looking complete:

• Implemented: operational CII (attained, required, A–E rating) for all 12 ship types, with coefficients verified against the IMO resolution PDFs; speed–consumption baseline fitting; clean-hull fouling-trend detection; flexible noon-report ingestion with surfaced data-quality flags.
• Not yet implemented: the CII correction factors permitted for certain ship types and voyages under MEPC.355(78) and the G5 guidelines — for example, voyage adjustments (ice transit, STS operations) and exclusions of electrical power consumed by reefer cargo or specialised systems. The current engine computes the uncorrected operational CII. For many ships this is the whole story; for ships eligible for corrections, the attained CII reported here will be conservative (i.e. it may look slightly worse than the corrected figure). This is on the roadmap and flagged here so results are never over-interpreted.
• Not a compliance authority. See the disclaimer below.

• CII correction factors and voyage adjustments (MEPC.355(78) / G5)
• EU MRV / IMO DCS report export
• EEXI (design index) calculation alongside operational CII
• Trim/draft correction in the performance baseline
• Multi-fuel voyage aggregation directly from per-leg noon data
• Optional local web dashboard (charts + uploads) on top of this library

Issues and PRs welcome — especially corrections traced to a primary IMO source. If a coefficient is wrong, cite the resolution and it gets fixed with a test.

MIT. See LICENSE.

This software is provided as-is. Results are estimates and guidance and must not be treated as proof of regulatory compliance. Always verify against official sources and your Administration / Recognized Organisation.

noonkit is built and maintained by Navallogic Solutions — an independent maritime advisory focused on vessel-performance analytics and decarbonisation compliance — as part of an open portfolio that includes the FuelEU Pool Optimiser (lowest-cost fleet compliance via linear programming) and the Maritime GHG Compliance Navigator (one fuel picture scored against IMO NZF, FuelEU, EU ETS and CII simultaneously).

For fleet-wide performance audits, hull-cleaning timing studies, or CII trajectory planning beyond what a public tool can responsibly offer, get in touch via navallogic.com.

Built by Rizwan Ali Mondal.