Ro-Pax EV-Deck Pilot: Detection Results

This pilot asked a question a pure car carrier never has to: how much earlier does per-vehicle detection flag a developing event on a vehicle deck that has people asleep above it? A short-sea ro-pax carrying passengers and drivers on overnight crossings ran a thermal-and-gas anomaly layer across a season, alongside its fitted detection, with every alarm logged on a common time base. On an occupied ferry the lead time is not asset margin — it is evacuation margin. [VERIFY: pilot is a representative composite pending a confirmed operator dataset — founder to verify or replace figures before publish.]

The vessel and the window

• ~2,100 lane-metre short-sea ro-pax, overnight crossings with ~550 passengers and accompanying drivers aboard.
• Roughly a 100-sailing season, mixed sea states, short turnarounds alongside.
• Vehicle mix around 22% battery-electric cars, balance ICE, plus vans and coaches.
• No onboard charging — consistent with EMSA guidance that charging is permitted only after a ship-specific risk analysis.
• Anomaly layer running alongside, not replacing, the ship's fitted smoke and heat detection.

Why a ro-pax is a different detection problem

Because the vehicle deck is occupied space, and on the overnight leg it is closed and unmanned while people sleep aboard. On a PCTC a detection layer buys salvage and firefighting margin; on a ferry it buys the time to warn and muster passengers and drivers before an enclosed deck fills with smoke — the exact failure mode that killed 11 truck drivers on the Euroferry Olympia (2022), found in their cabs on the vehicle decks. EMSA's guidance on carrying alternative-fuel vehicles in ro-ro spaces (published April 2025, updated November 2025) stresses stowage that keeps every vehicle accessible to the crew; and amended SOLAS Chapter II-2 Regulation 20, via MSC.421(98), now requires ro-pax weather decks intended for vehicles to carry fire detection and water-based monitors on new-builds from 1 January 2026. The pilot deployed to both the enclosed vehicle deck and the weather deck for that reason.

Deployment: non-invasive retrofit on a ferry schedule

The constraint a ferry adds is time alongside. A ro-pax turns around in hours, not days, and cannot be taken off a fixed timetable for a structural refit. Installation ran as a non-invasive retrofit with zero hull modification, staged across several nights alongside during normal turnarounds — no hot work, no drydock, no class-society deviation, and no cancelled sailings. The point of proving the retrofit path on a ferry is that the timetable, not the engineering, is usually the blocker; a layer that installs between departures is one an operator can actually adopt.

"On the night crossing the car deck is shut and everyone is up in the cabins. The console gave us a vehicle and a position while the deck was still clear enough to walk. That is the difference between a walk and an evacuation."

— Chief Officer (operator NDA) [VERIFY]

What needed tuning

The ro-pax environment moved the baseline in ways a car carrier does not. Coaches and refrigerated vans left running until the last minute put localised engine and exhaust heat on the deck that the per-vehicle baseline had to learn and discount; heavy passenger and rider foot traffic during loading added transient thermal clutter; and salt, humidity and the weather deck's sun-loading and wind widened the background further. The first stretch of the season ran in calibration while the magnitude and dwell thresholds settled against that profile. Once the coherence window was tuned to the ship's ventilation and loading pattern, the false-alarm count held at zero for the remainder — which on an occupied ship matters doubly, because a nuisance alarm that trains a crew to hesitate is itself a life-safety cost.

Why it matters to the operator

Four pre-smoke early-warnings in a season, each locating a specific vehicle around 18 minutes ahead of the ship's fitted detection and delivered to the bridge while the overnight deck was unmanned, is the operational case for the layer on a ferry: it converts a no-warning event into a window in which the crew can verify, isolate and — if needed — begin mustering before the enclosed deck becomes impassable. That is the same margin the Euroferry Olympia did not have. It also sits with the regulatory direction: as SOLAS pushes detection onto ro-pax weather decks and EMSA presses for crew access and charging discipline, an early per-vehicle layer is how an existing ferry closes the gap without leaving the water.

Sources

• RoRoSAFE ro-pax pilot record (operator under NDA) — ~2,100 lane-metre short-sea overnight ferry, ~100-sailing season, main vehicle deck + weather deck, ~22% BEV car mix; 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; the early-warning count, 18-minute lead, per-vehicle-to-bridge integration and zero false-alarm figures are illustrative, not measured field results.]
• EMSA — Guidance on the Carriage of Alternatively Fuelled Vehicles (AFVs) in Ro-Ro Spaces (published 28 April 2025, updated 24 November 2025): stow so crew retain access to all vehicles; onboard charging only after a ship-specific risk analysis — emsa.europa.eu.
• IMO — SOLAS Chapter II-2 Regulation 20 as amended by MSC.421(98): ro-pax weather decks intended for vehicle carriage to be fitted with fire detection and water-based fire monitors; new-build application from 1 January 2026 — imo.org. [VERIFY: confirm the MSC resolution number and the weather-deck detection/monitor wording against the SOLAS text before publish.]
• EMSA — FIRESAFE I and FIRESAFE II studies of fire dynamics and system performance in ro-ro spaces — the evidence base for ro-ro/ro-pax vehicle-deck fire safety — emsa.europa.eu.
• Companion RoRoSAFE analysis — 'Euroferry Olympia: When the Deck Had People' (the ro-pax life-safety anatomy this pilot is measured against), 'Should EVs Charge on a Passenger Ferry?' (the charging-discipline question), and 'PCTC Retrofit Pilot: Detection Results' (the car-carrier sibling pilot).