Shipping's compliance burden is tightening across all operational levels

The Genbu notation, granted under ClassNK's autonomous vessel guidelines, authorises remote-controlled coastal operation without any onboard crew. The vessel uses an Action Planning and Execution System (APES) for navigation and course-correction, with all machinery monitored and adjusted by a shore-based control centre. This represents IMO MASS Degree 3 operation — the highest current autonomy level in commercial service.

NYK's Elder Leader, delivered for deep-sea car carrier service, represents the contrasting approach. During Singapore Maritime Week pilotage trials, the vessel integrated its route plans directly with MPA's Next-Generation Vessel Traffic System in real time — demonstrating highly automated, shore-linked operation on international routes. However, the vessel operates at IMO MASS Degree 2: highly automated with a manned bridge and engine room under STCW-compliant manning requirements.

For watchkeeping engineers on deep-sea vessels, the message from this comparison is operationally important: AI routing and automated performance monitoring are now standard equipment, not premium features. The engineer's role on a Degree 2 vessel is to validate AI output, manage exception cases, and retain the diagnostic capability that cannot yet be automated — complex machinery diagnosis, fuel system adjustments, and emergency response.

The anxiety around replacement by automation is not supported by current commercial deployment patterns. Degree 3 autonomy remains restricted to short coastal routes with predictable traffic and benign weather. Deep-sea merchant operation — variable weather, complex machinery, port-state requirements, and international watchkeeping conventions — sustains the demand for onboard engineering professionals for the foreseeable operational horizon.

MEPC 84, the 84th session of the IMO Marine Environment Protection Committee, concluded with geopolitical division preventing adoption of the binding Net-Zero Framework (NZF), including the proposed GHG fuel standard and global economic pricing mechanism. Negotiations transfer to intersessional working groups ahead of MEPC 85. The decision defers the carbon pricing mechanism but does not pause the operational audit regime that is already in force.

Two CII-related decisions at MEPC 84 carry direct onboard implications. First, the committee confirmed that the Phase 2 review of the Carbon Intensity Indicator — effective from 2027 — will retain the current calculation boundary: total fuel-based emissions divided by total distance sailed. Proposals to exclude port and anchorage periods from the denominator were rejected, maintaining the incentive for voyage efficiency over the full operating profile rather than underway segments only.

Second, the committee discussed a proposed revision to MEPC.1/Circ.905 that would align the carbon intensity factor of biofuel blends toward a mass-weighted average of the blend components. The current energy-weighted method produces different carbon factors for the same physical blend depending on calculation approach — the proposed change would align the calculation with actual mass composition and close an inconsistency in biofuel reporting across flag states. Engineers should monitor the formal circular revision for confirmed implementation dates.

For Port State Control purposes, SEEMP Part III logs are subject to increasing scrutiny under MARPOL Annex VI as the primary enforcement tool while the NZF negotiations continue. The practical checklist for Chief Engineers has not shortened: fuel figures must be accurate, distance records must be reconciled against ECDIS logs, and SEEMP execution must be documented at the frequency specified in the approved plan.

IACS UR M78 Revision 4 formalises what the decarbonisation agenda has made inevitable: ammonia is now a machinery-space engineering subject, not a future-fuels discussion topic. The revision applies to all IACS member classification societies — Bureau Veritas, ClassNK, DNV, Lloyd's Register, RINA, and others — and mandates specific design requirements for any vessel fitted with an NH₃-fuelled reciprocating internal combustion engine.

Ammonia presents two distinct and simultaneous hazards. As a toxic gas, NH₃ has a threshold limit value (TLV) of 25 ppm over an 8-hour exposure, requiring continuous monitoring due to both its toxicity and corrosive properties. As a corrosive substance, it degrades copper alloys and zinc commonly found in heat exchangers, instrumentation lines, and electrical components. The UR M78 design requirements address both hazard pathways.

For MEO Class 1 oral examination purposes, IACS UR M78 Revision 4 now constitutes primary source material for any question on alternative fuel machinery design. Examiners may ask candidates to describe the double-walled piping philosophy, compare NH₃ and LNG machinery-space requirements, or explain the rationale for segregated fuel preparation rooms.

In Issue 18, Google Analytics 4 (GA4) was deployed across this publication to track section-level reader engagement. The stated objective: apply Andrew Ng's AI feedback loop methodology¹ to editorial production — measure what readers actually consume, compare against editorial intent, and adjust the product accordingly. The period from 7 April to 4 May 2026 provided the first usable dataset.

The data contains one apparent anomaly. In engineering diagnostic terms, two sensors return contradictory readings simultaneously. This indicates rapid scrolling behaviour rather than sustained reading — users reach the end of the publication quickly without pausing on individual sections, triggering the depth event without dwell time. This is a structural entry problem, not a content-quality problem, and it warrants a structural correction.

The combination of 13-second average duration and 83% scroll-depth completion produces a specific diagnostic: readers are fast-scrolling through the publication without pausing to read, but they are not leaving early. This is the behaviour of a reader who is scanning for an anchor — a headline, a statistic, or a section format that compels them to stop and read rather than continue scrolling.

The cover-entry problem is structural. A reader arriving cold encounters an abstract headline label ("MEPC 84 Outcomes"), a dense text panel, and a multi-item navigation bar before encountering the Quick Read cards that actually arrest attention. Issue 19 addresses this directly: the cover headline restates the news angle as a consequence rather than a label, the one-paragraph summary appears on the cover itself, and the editorial disclaimer has moved to the author page.

The 16.7% return visitor rate — below the 30% target — indicates that readers who visit once are not returning for a second issue. The most likely cause is insufficient cover gravity on the first visit: if the entry experience does not immediately signal that the publication's content is directly relevant to the reader's professional concerns, the probability of a second visit decreases. This is the next feedback loop variable to resolve.