The vessel's condition record, per hull, day by day

A record that builds against the hull

LOOP runs aboard on Android tablets, a Windows hub and a Raspberry Pi engine-room display, with the database and the AI assistant on the ship itself. No connectivity is required for any core function, which is why the daily uplink is a few kilobytes rather than a line on a per-GB Starlink bill. What lands ashore each day is the hull's own condition: rounds compliance, worklist state, alarm state, headline trends with 30-day series, plant changeovers, recent notes and compressed issue photos.

• Roughly two documents per vessel per day, typically under 5 KB.
• Rounds keep running and the shipboard record keeps building with no connection at all. The uplink only decides what shore sees, not what the ship captures.
• About 20 headline trends, each with a 30-day series and a count of 2-sigma anomalies, alongside rounds compliance for today and the last 7 days.
• Up to eight compressed issue photos a day, shown as a grid on the vessel page ashore.
• The shipboard record layer is mutable and capacity-capped today. Append-only, signed, tamper-evident storage is next, and until it lands this is a condition record, not an evidence trail. in build

Consumption rates, trended per hull

Consumption is computed aboard from the readings the watch already takes: fuel in tonnes per day, cylinder lube oil in litres per day. The computed-metrics engine handles deltas, rates per hour and per day, two-element expressions and unit conversion, so the number in the rollup is derived at the gauge rather than reconstructed ashore weeks later. Each vessel page charts this ship against the class. The class line is only as good as the hulls reporting into it, and today that is one.

• Fuel tonnes per day and cylinder lube oil litres per day, per hull, per day.
• Inline alarm evaluation against configurable conditions, so an out-of-range reading is caught as it is entered, not at month end.
• This-ship-vs-class charts on each vessel page, plus a fleet league table and rounds and worklist comparison tables ashore.
• Live AIS positions for the fleet on a weather map, with 24-hour dead-reckoned tracks. Three of the five hulls have been seen reporting real positions.
• Sister-ship deviation bands need at least three hulls on real data; one reports today, so no deviation figure is published against the class. in build

Readiness, and what the number is not

The vessel page carries a readiness dial, 0 to 100, with the component breakdown that produced it, so the number is never presented on its own. The dashboard carries stakeholder lenses, and an owner's reading of a dragging component is a different question from a manager's: parts never ordered is not the same failure as rounds not walked. The breakdown is what lets you tell those apart, and whose line item it is stays a conversation in the room rather than a verdict in the software. The weighting is ours, not a class rule or a BIMCO standard, and it should be read as an indicator, not a grade.

• The component breakdown sits next to the dial, not just the headline number.
• A component dragging because parts were never ordered is a different failure from one dragging because rounds were missed. The breakdown is what tells you which.
• A priority "needs attention" panel lists what is actually open, so the score is never the only thing on the page.
• No individual scoring, no stopwatch metrics, no surveillance. The crew sees what shore sees. This is a union-crewed US-flag fleet and the tool is not used for appraisal.
• There is no per-user login aboard yet, so per-reading attribution is not meaningful and is not presented as if it were. Per-user named authentication is next. in build

About money, plainly

There is no measured saving in this product, on any hull, and there will not be an invented one. Any dollar figure is modelled, with the assumption and the date printed beside it. Cross-hull deviation, and any money framing built on it, stays off the dashboard until at least three hulls are sending real rollups — with one real meter and four sets of demo data, a deviation figure would be an artefact of the demo, not a finding about a ship.

• No ROI, payback or savings figure appears in the product or on this site.
• Modelled illustrations carry their assumption and their date on the same card. If the assumption is not shown, the figure is not shown.
• Demo data is labelled as demo wherever it appears, the league table and comparison tables included, so a mixed-source number is not read as a result.
• Cross-hull deviation waits for three or more hulls on real data. Today it is withheld.
• What a pilot costs is a conversation with the manager in the room, not a number derived from savings nobody has measured.

Roadmap for the owner side

These are US-flag hulls, and some of what follows is specific to that flag: MSP operating-agreement and DoD availability, and CARB at-berth in California. The owner-facing work is queued behind the record store, because none of it is worth much on a mutable record. Everything in this section is roadmap. None of it runs today.

• CII/AER shown as an estimate and labelled as an estimate. The IMO DCS filing and the verifier relationship stay with the manager. in build
• CARB at-berth evidence pack, assembled from what the rounds already capture. in build
• Off-hire hours and critical-failure counts as reliability KPIs. Needs new fields in the ship rollup before the numbers mean anything. in build
• MSP/DoD readiness file for US-flag MSP hulls. in build
• Defect lifecycle with a shore-acknowledgement field and an exportable open-defect register, plus true-copy PDF exports for survey and incident bundles. in build

What is shipped, and what is not

This table is the point of the page. It is written to be read by a sceptic, and it is kept honest deliberately — every claim here is checkable against the running system.