Can an Autonomous Ship Fight a Deck Fire?

Not the way a crewed one does. The IMO adopted its first Code for Maritime Autonomous Surface Ships at MSC 111 in May 2026, and it took effect on 1 July 2026 — non-mandatory, for cargo ships, with chapters covering both manning and fire protection, detection and extinction. On a vehicle deck the crew is not an accessory to the fire plan; it is most of the plan. Remove or relocate it and detection has to carry weight it was never designed to carry.

What the MASS Code actually did

It created the first global framework for ships operating with reduced, remote or absent crews — and started the clock on a mandatory version. Adopted at the Maritime Safety Committee's 111th session (13–22 May 2026) and in effect from 1 July 2026, the Code applies to cargo ships under SOLAS Chapter I and is non-mandatory for at least an initial period, giving flag States an experience-building phase before it hardens. Its chapter list is broad — approval process, risk assessment, operational context, system design, software principles, alert management, manning, training and watchkeeping, remote operations, and fire protection, fire detection and fire extinction among them. Two features matter most here. The Code integrates Remote Operations Centres as a recognised part of the structure, and it holds that the master retains overall responsibility for the ship even when not on board. The IMO's stated path is a mandatory Code adopted by around 1 July 2030 and entering force on 1 January 2032.

Why a vehicle deck is the hardest case

Because on a car deck the crew performs functions no fixed system performs, and they are exactly the functions that disappear first. When a detector trips on a PCTC today, a person goes and looks. They confirm what is burning and where, close the fire boundaries, verify that the space is clear before a drencher or CO2 release, start boundary cooling on adjacent decks, and keep re-assessing as the event develops. None of that is sensing — it is judgement and physical intervention among thousands of tightly stowed vehicles on a large undivided deck. A remote operator ashore inherits the judgement and loses the intervention. That is a different problem from navigation autonomy, where the machine can do the task; here the machine cannot walk the deck, and the fixed installation cannot decide.

Verification is the function that vanishes first

The single most human part of car-deck fire response is deciding whether the alarm is real, and that is the part a remote centre cannot do. Nuisance alarms on vehicle decks are a known operational cost, and today they are resolved by someone walking to the source and looking. Take that away and the discrimination has to happen in the sensing layer itself, because a Remote Operations Centre cannot investigate — it can only act or not act. That raises the bar in both directions at once: a missed event has no human backstop, and a false one may trigger an irreversible fixed-system discharge on a full cargo deck. Detection quality stops being a question of lead time and becomes a question of whether the evidence is good enough to commit a decision no one can walk back.

'Fixed First' stops being a preference

IUMI's fixed-first posture for vehicle carriers is already the direction of travel; on a reduced-crew ship it is not a preference but the entire plan. If the fixed installation is the only actor, then the trigger decision has to be made on sensor evidence alone — and that changes what the sensing has to deliver. Per-vehicle localisation is the clearest example. 'There is smoke on Deck 5' is actionable for a crew who can go and find it; for an operator ashore weighing a drencher or CO2 release across a whole deck, it is not enough information to act well. The reduced-crew case therefore pushes detection toward the same properties a good crewed system wants — early, specific, and locating the source vehicle — but removes the human margin that currently forgives their absence.

What it means for owners, class and underwriters

• Treat the fire chapter as the binding constraint on autonomy for vehicle carriers. Navigation autonomy has a machine substitute; car-deck fire response largely does not, so fire response is where a reduced-crew PCTC case will be won or lost in approval.
• Expect class and flag to probe the degraded-state fallback: what happens when detection is ambiguous, connectivity to the Remote Operations Centre is lost, or the fixed system has already discharged once.
• The master retains responsibility even when ashore — so the evidentiary record of what the detection system saw, and when, becomes central to accountability rather than incidental to it.
• For underwriters: reduced crewing on a vehicle carrier changes the loss model, not just the crew cost. Ask what replaces the human verification step before pricing an autonomy discount.
• The experience-building phase to a mandatory Code around 2030–2032 is the window to establish what adequate vehicle-deck detection looks like on a reduced-crew ship — before the requirement is written.

Sources

• IMO — 'IMO adopts first global Code for autonomous ships' (press briefing): the non-mandatory MASS Code was adopted at MSC 111 (13–22 May 2026) and took effect on 1 July 2026; it applies to cargo ships under SOLAS Chapter I; it integrates Remote Operations Centres, and the master retains overall responsibility for the ship even if not on board; expected adoption of a mandatory Code by around 1 July 2030 with entry into force 1 January 2032 — imo.org. [VERIFY: the adopting resolution is reported elsewhere as MSC.595(111) but was not confirmed on IMO's own page; confirm the resolution number before publish.]
• MASS Code chapter scope, as reported from the draft: chapters covering surveys and certificates, approval process, risk assessment, operational context, system design, software principles, management of safe operations, alert management, manning/training/watchkeeping, safety of navigation, connectivity, remote operations, stability, fire protection/fire detection/fire extinction, security, search and rescue, carriage of cargoes, mooring, and machinery — Maritime Executive / Baird Maritime / gCaptain coverage of MSC 111. [VERIFY: chapter list is from trade coverage of the draft; confirm the adopted structure against the Code text before publish.]
• Voluntary application for an initial period (reported as at least two years) to allow Member States an experience-building phase before the Code is made mandatory under SOLAS — trade coverage of MSC 111. [VERIFY: confirm the stated duration of the experience-building phase.]
• Unmanned and reduced-crew operation requires designing for emergency response without a physical crew to fight fires or manage localised failures, with fallback for 'reasonably foreseeable degraded states' — trade reporting on the MASS Code's operational-context and risk-assessment provisions. [VERIFY: confirm this wording against the adopted Code.]
• IUMI — position and best-practice material advocating a 'Fixed First' firefighting approach for vehicle carriers, and identifying the limits of manual intervention on vehicle decks — iumi.com.
• Companion RoRoSAFE analysis — 'Who Fights a Mid-Ocean Car-Carrier Fire?' (response reach on a crewed ship), 'IUMI 2025: The Fixed-First Approach' (the doctrine this post extends), and 'The Cost of a False Alarm at Sea' (why verification is the expensive step).