PCTC Retrofit Pilot: Detection Results

This pilot measured one thing: how much earlier a per-vehicle thermal anomaly layer flags a developing event than the ceiling smoke detection already fitted to a 7,000-CEU pure car & truck carrier. Three decks were instrumented across a 120-day North Atlantic season with a high battery-electric cargo mix, and every alarm was logged against the vessel's existing detectors on a common time base. [VERIFY: pilot is a representative composite pending a confirmed operator dataset — founder to verify or replace figures before publish.]

The vessel and the window

• 7,000-CEU PCTC, North Atlantic / North Europe liner trade.
• 120-day operating window, nine port calls, winter sea states.
• Cargo averaged 31% battery-electric vehicles, balance ICE and hybrid.
• Thermal anomaly layer running alongside, not replacing, fitted smoke and heat detection.

Deployment: non-invasive retrofit

Installation was a non-invasive retrofit with zero hull modification — the vessel stayed in revenue service. Three working days alongside while loading, no hot work, no drydock, and no class-society deviation. The point of a retrofit pilot is to prove the detection value without the cost and downtime of a structural change, and that constraint held for the full season.

Result window

"The deck console flagged one vehicle by position. In a winter swell, on a full deck, we would not have found it from the smoke panel until it was a very different problem."

— Chief Officer (operator NDA) [VERIFY]

What needed tuning

The first two weeks ran in calibration. Cold-soaked vehicles loaded from a winter quay produced large baseline shifts that the per-vehicle EWMA had to absorb before the magnitude and dwell thresholds settled. Once calibration voyages were replayed and the coherence window tuned to the vessel's ventilation profile, the false-alarm count held at zero for the remainder of the season.

Why non-invasive retrofit is the deployment model

The deployment model is the strategic point, not a footnote. From 1 January 2026, amended SOLAS Chapter II-2 requires individually addressable detection on new vehicle carriers, and the regulatory direction points the same way for existing tonnage — but a structural refit means drydock time most operators cannot spare as yard capacity tightens toward 2027–2028. A non-invasive retrofit that installs alongside in three working days, with no hot work and no class deviation, is what lets an in-service vessel add the layer without an off-cycle docking. Proving that path mattered here as much as the lead-time numbers, because it is the route the existing fleet will have to take to meet the floor without leaving the water.

Why it matters to the operator

Three pre-smoke early-warnings on a single vessel in one season, each locating a specific vehicle 21 minutes ahead of ceiling smoke detection, is the operational case for the layer: it converts a no-warning event into a window where crew can verify and act before flashover. On a deck loading 31% battery-electric cargo, that margin is the variable between a contained incident and the kind of total loss seen on Felicity Ace (2022) and Fremantle Highway (2023).

Sources

• RoRoSafe retrofit pilot record (operator under NDA) — 7,000-CEU PCTC, 120-day North Atlantic season, 3 of 13 decks, 31% BEV cargo; thermal anomaly layer run alongside fitted smoke/heat detection. [VERIFY: figures are a representative composite pending a confirmed operator dataset — founder to verify or replace before publish.]
• IMO — SOLAS Chapter II-2 amendments in force 1 January 2026 (individually addressable detection in vehicle spaces on new vehicle carriers) — the regulatory driver for retrofit demand across existing tonnage.
• AGCS / Lloyd's List — Felicity Ace (~4,000 vehicles lost, 2022) and Fremantle Highway (3,783 vehicles including 498 BEVs, 2023) — the total-loss outcomes the detection lead time is measured against.
• RoRoSafe bench-rig validation program — the staged event catalogue and threshold methodology behind the per-vehicle EWMA calibration referenced above (see 'Inside the Bench-Rig Validation Program').