Japan's MEGURI2040: What Level 4 at Sea Actually Looks Like

Week 13 | March 2026

While US Coast Guard lawyers were still parsing a statute written in 1987, a 66-metre Ro-Pax ferry carrying 500 passengers threaded through one of the world’s most congested shipping corridors without a hand on the wheel.

No incident. No press conference from a congressional committee asking pointed questions about who is responsible. Just a ferry, moving people across the Seto Inland Sea, doing what it was built to do.

Japan didn’t wait for the IMO. They built an institution instead.

The vessel is a 942 GT Ro-Pax ferry, retrofitted in three stages through mid-2025, operating the Shin-Okayama to Tonosho (Shodoshima Island) route. The crossing takes roughly 70 minutes through the Seto Inland Sea, with traffic including fishing vessels, cargo ships, and debris. On December 5, 2025, it became the first vessel in the world to receive Japan’s formal “autonomous ship” designation under safety standards published by the Ministry of Land, Infrastructure, Transport and Tourism. Commercial operations at full deployment are scheduled for end of March 2026.

The system stack is genuinely sophisticated. HD cameras and LiDAR generate continuous 3D environmental maps. An AI navigation layer predicts the position of nearby vessels at 5, 10, and 30-minute intervals and adjusts course to maintain COLREGS-compliant passing distances. Autonomous berthing and unberthing from narrow wharves is included. Three independent collision avoidance systems, developed separately by Mitsubishi Shipbuilding, Mitsui E&S Shipbuilding, and Japan Marine Science, were run in parallel and mutually refined — a redundancy approach specific to this program. Shore-based communications run on satellite and 5G with military-grade encryption, routed through a permanent Fleet Operation Center in Nishinomiya City.

Ten crew remain aboard during operations. Their role, as one project official put it, is essentially emergency response: evacuation and override if something goes fundamentally wrong. A shore-based master mariner monitors by satellite. The vessel operates under a government exemption from the COLREGS Rule 5 requirement for a human lookout, which is the legal mechanism that made certification possible.

“Level 4” in this context means the vessel handles navigation and collision avoidance autonomously within defined operational conditions without requiring human intervention. It is analogous to, but not formally derived from, the automotive SAE framework. The IMO MASS Code uses different terminology entirely. What matters is the operational reality: this vessel is navigating one of Japan’s most congested waterways, with passengers aboard, autonomously, in commercial service.

Japan’s approach rests on a trilateral model that no AUKUS nation has attempted to assemble.

The Nippon Foundation, Asia’s largest private grantmaker with roughly $600 million in annual philanthropic spending, serves as the primary funder and organiser for MEGURI2040. The Ministry of Land, Infrastructure, Transport and Tourism provides regulatory enablement, publishes safety standards, and chairs the navigation chapter of the IMO MASS Code working group. ClassNK, the classification society, is embedded from the beginning as both standards developer and risk assessor, creating a direct pipeline from demonstration to certification. The industry consortium, now at 51 to 53 companies including NYK, MOL, K-Line, Mitsubishi Shipbuilding, and Furuno Electric, develops and deploys the technology.

The funding scale, long unclear in English-language sources, is now verifiable. Stage 1 of MEGURI2040 cost approximately 8.8 billion yen in total, of which the Nippon Foundation subsidised roughly 7.4 billion yen directly. Stage 2, running from late 2022 through March 2026, carries a technical development grant budget of 10 billion yen. Combined with additional safety-evaluation grants, known MEGURI2040 spending across both stages is at least 18 to 19 billion yen, with the Foundation directly subsidising around 17 to 18 billion yen of that total. Call it $120 to $130 million USD at current exchange rates. For context, the US has zero civilian federal investment in autonomous commercial shipping. Zero. The Navy has programmed $3 to 4 billion for military autonomous vessels. The commercial sector waits.

The legislative foundation is equally important and gets almost no attention outside Japan. In 2021, the Act on Strengthening Maritime Industries amended six separate laws simultaneously: the Marine Transportation Law, Shipbuilding Law, Coastal Shipping Law, Mariners Act, Seafarer’s Employment Security Act, and Ship Safety Law. One omnibus act, six statutes, one legislative session. The US equivalent would require Congress to separately amend each statute, and the Coast Guard has explicitly told the GAO it lacks authority to waive statutory crew minimums under the current framework, outside a single narrow pilot program for SpaceX rocket recovery.

This is the institutional gap. Not technology. Not money, exactly. But the ability to align a foundation, a ministry, a classification society, and 50-plus companies around a shared phased roadmap, then change six laws in one go to clear the regulatory path.

Phase 2 of MEGURI2040 targeted four vessels for Level 4 commercial certification by April 2026. Three are cargo ships.

The Genbu is the standout. A new-build 696 TEU container vessel, delivered October 2025 and designed from the keel up for autonomous operations, it passed Japan’s national ship inspection as an “autonomous operation ship” on January 28, 2026, and entered commercial service on the Kobe–Osaka–Nagoya–Yokohama route on January 30. Per Nippon Foundation documentation, this constitutes the world’s first Level 4-equivalent commercial autonomous operations on a container ship. Not a demonstration. Not a trial. Commercial service.

Two other vessels, Mikage and Hokuren Maru No. 2, are a different story. Both successfully completed proof-of-concept demonstrations in earlier phases: Mikage ran a 161-nautical-mile autonomous transit in 2022, and Hokuren Maru No. 2 completed three autonomous voyages on a 1,600-kilometre round trip in 2023 with a 96 percent system operation rate. Both began Phase 2 certification demonstrations in mid-2025, targeting April 2026 certification. As of publication, neither has publicly confirmed passage of government inspection as an autonomous ship or entry into Level 4 commercial operations. The April deadline is close.

On February 21, 2026, the MV Zhi Fei completed what multiple sources are calling the world’s first end-to-end fully autonomous commercial container vessel operation at Qingdao Port’s New Qianwan Container Terminal.

The integrated chain: autonomous navigation, autonomous docking and berthing, vacuum-based automated mooring secured in under 30 seconds with no mooring lines, and automated cargo handling via intelligent quay cranes and AGVs with millisecond-level response. No human intervention across the full sequence from ship arrival to cargo transfer.

This is not a new vessel. The Zhi Fei has been in commercial service since April 2022, operating the Qingdao–Dongjiakou route. By February 2026, it had completed 353 voyages in 2025 alone, transported 80,800-plus TEU, logged 48,000-plus nautical miles since service began, and recorded over one million independent AI decisions. The February milestone is not about the vessel proving itself. It is about integrating the ship-to-port autonomous chain into a single unbroken operation at commercial scale.

BRINAV, the operator, developed the navigation AI domestically. Export controls on foreign radar brands accelerated indigenous development. The communications stack runs on 5G connectivity optimised by all three major Chinese carriers simultaneously. The port-side automation runs on domestically developed terminal operating systems.

The dual-use implications deserve their own edition, and they will get one. For now, the same autonomous docking capability demonstrated by the Zhi Fei has direct application to unmanned resupply operations. Domestically developed radar, communications, and AI navigation systems are directly transferable to military autonomous platforms. China’s national defence mobilisation laws and technical standards requiring reinforced decks and stronger ramps on major civilian vessels formalise the pipeline from commercial deployment to military utility. The Zhu Hai Yun autonomous drone mothership (built by entities that are also in the Zhi Fei supply chain ecosystem) and its unauthorised entry into the Philippine EEZ last July give that pipeline operational context.

Yara Birkeland gets cited in almost every autonomous shipping piece as proof the technology works in commercial operations. The picture is more complicated.

The vessel has completed 250-plus voyages, transported 35,000-plus containers on the Herøya–Brevik route since April 2022, replaced roughly 35,000 diesel truck journeys, and reduced CO₂ by around 1,000 tonnes per year. Auto-docking and auto-crossing are operational. Crew has been reduced from five to three, with a further reduction to two imminent.

But full autonomous certification, the point at which Yara Birkeland would operate without any crew aboard, was originally targeted for the end of 2024. Then early 2026. As of publication, the Norwegian Maritime Authority has not issued a final certificate for fully unmanned autonomous operations. The autonomy trial is still running.

This is not a failure. Norway is operating at the frontier of what international maritime law currently permits, moving carefully to ensure the regulatory precedent holds. Their Maritime Authority has a dedicated seven-person department for new maritime technology and has adopted what they describe as a “partner in development” approach to the industry. In May 2024, Norway, the UK, Belgium, Denmark, and the Netherlands formed the North Sea Autonomous Shipping Alliance to standardise requirements pending IMO regulations.

But the headline “Yara Birkeland is fully autonomous” is not accurate. Not yet. Japan’s Genbu entered Level 4-equivalent commercial service two months ago. Norway is still working through the certification process for unmanned operations that Japan has now completed for crewed-aboard autonomous operations.

The non-mandatory MASS Code has not been adopted as of publication. At MSC 110 in June 2025, 18 of approximately 26 chapters were finalised. MSC 111, scheduled for May 13 to 22, 2026, is the target adoption date. The mandatory code follows: development begins in 2028, adoption targeted July 2030, entry into force January 2032.

A key provision worth noting: under the emerging framework, there must always be a human master responsible for a vessel regardless of operational mode, though the master need not be physically aboard. Unmanned vessels must be capable of assisting persons in distress at sea. These are not trivial requirements, and they will shape commercial deployment economics significantly when they become mandatory.

The practical point: Japan, China, South Korea, and Norway are all operating under flag-state authority without waiting for the international framework to be established. South Korea’s Avikus has now surpassed 100 cumulative autonomous navigation system installations on large commercial vessels, including a 40-vessel contract with HMM signed in January 2026, the largest single autonomous navigation contract ever. The framework will eventually matter. Right now, it is largely irrelevant to the countries actually deploying.

The US paradox runs deep. The Navy is building autonomous warships: $3 to 4 billion programmed for LUSV and MUSV programs, Sea Hunter and Sea Hawk completing a seven-month Pacific deployment of 46,651 nautical miles in near-constant autonomous mode, DARPA’s USX-1 Defiant commissioned in August 2025. The commercial fleet has nothing. Coast Guard regulations were last updated for automated vessels in 1987 and 1988. The agency has told Congress it is waiting for the IMO framework before issuing domestic guidance, meaning US commercial maritime regulations will lag even the non-mandatory code, due in May 2026.

Japan built a foundation-ministry-industry institution and changed six laws at once. South Korea programmed $167 million in confirmed civilian autonomous shipping investment with year-over-year increases. China is integrating autonomous commercial operations into a documented military-civil fusion strategy. Norway has a seven-person regulator team that partners with industry rather than waiting for international consensus.

AUKUS nations, Australia included, have the military side developing at pace. Ghost Shark is in production. The civilian regulatory parallel is nowhere. Week 3’s Coast Guard piece covered the US dimension. The Australian equivalent - what statutory barriers apply to autonomous commercial vessels in Australian waters, how the Navigation Act 2012 intersects with the IMO MASS framework, and whether any regulatory sandbox equivalent to Japan’s exists- is slated for future OTI coverage.

The institutional design question is ultimately more important than the technology question. The technology works. Japan just demonstrated that in commercial service, with passengers aboard. What the AUKUS nations lack is not engineering capability. It is the institutional will to change multiple laws at once.

We shift to offshore wind’s automation revolution, and why the workforce crisis in European offshore wind is creating the commercial case for autonomous crew transfer vessels that the maritime sector has been waiting for.

Since you have been, thanks for reading.

Cheers,

Mick

Ocean Tech Intelligence provides informational analysis only. Nothing in this publication constitutes financial, investment, legal, or strategic advice. Readers act on this content at their own risk. For full details, see our Disclaimer.