Nixon V Antony
Second Engineer · Container Vessels · Maersk A/S · Editor, Marine Intelligence Weekly
Japan has certified four autonomous vessels. Norway has formalised shore-based operator licences. Singapore has committed SGD 100 million to AI port infrastructure. India is beginning to build its own layer — through startups, smart ports, and operational intelligence.
The next major shift in shipping may not begin with a new fuel type or engine design.
It may begin inside operational dashboards, camera systems, predictive analytics platforms, and AI-assisted decision-support layers — quietly, at the margins of what engineers already do.
Across global shipping, maritime AI is moving from experimentation toward operational deployment. Japan is advancing autonomous navigation at a systems level. Norway has formalised the certification of shore-based remote operators. Singapore has committed over SGD 100 million to maritime AI research deployment. Large shipping companies are embedding AI into voyage optimization, emissions management, and predictive engineering.
India remains early in this transition. But the signals are becoming increasingly clear.
Unlike Europe or East Asia, India's maritime AI layer is not currently driven by autonomous vessels or state-led smart-shipping programs at scale. Instead, early movement is emerging through deep-tech startups, port digitization milestones, government-backed logistics AI initiatives, and a research ecosystem anchored at institutions like IIT Madras.
The focus is not on futuristic marketing language. The focus is on practical friction points: delayed visibility, repetitive documentation workflows, cargo movement inefficiencies, safety monitoring gaps, and the cost burden of disconnected operational data.
Shipping rarely changes through press releases. It changes through systems that quietly reduce friction — and through engineers who understand both the machinery and the data layer above it.
The onboard reality, however, remains far more conservative than the ecosystem developing ashore. Marine engineers still depend on practical judgement, machinery awareness, troubleshooting logic, and disciplined verification. AI systems may identify patterns. They do not replace engineering accountability.
This edition examines maritime AI across three layers: global frontline developments, India's emerging ecosystem in depth, and the onboard implications for marine engineers working through this transition.
Japan is building the ecosystem for autonomous fleet operations. Norway is certifying the people to run them from shore. Singapore is digitizing the ports they call at.
Japan's approach to maritime AI is distinctive: it is not building individual smart ships. It is building the ecosystem required to operate them at scale — vessel autonomy, shore-based fleet control, and AI-assisted engineering management, developed in parallel. The Nippon Foundation's MEGURI2040 program provides the umbrella structure, but each of Japan's Big Three shipping companies is operating distinct AI frameworks.
Operational focus: Integrated ship-to-shore autonomous ecosystem
Operational focus: Predictive maintenance, route optimization, fleet analytics
Operational focus: Engine-room AI, retrofit automation, CBM analytics
Norway's autonomous maritime leadership is built around close cooperation between technology companies, classification bodies, the Norwegian Maritime Authority (NMA), and research institutions. The regulatory environment is actively enabling deployment rather than waiting for international consensus.
Unlike Japan's autonomy focus or Norway's coastal uncrewed shipping programs, Singapore approaches maritime AI primarily through port efficiency, logistics data integration, and structured commercial AI adoption.
India's maritime AI activity is not primarily about autonomous ships. It is about cutting logistics costs, improving port turnaround, and applying AI to operational problems with clear commercial justification today.
India's maritime AI ecosystem is early. But something substantive is forming — and its character is distinctly different from the autonomous-vessel programs driving Japan and Norway.
The government is creating the infrastructure layer. Startups are building the operational intelligence above it. Research institutions are developing the technical foundations underneath it. That is not a weakness relative to Japan or Norway. It may be exactly the right sequencing for an economy scaling at India's pace.
In February 2026, VOC Port became the first major Indian port to deploy a full-scale Digital Twin platform. By integrating IoT sensors, LiDAR mapping, drone imaging, and CCTV networks, the system creates a live virtual replica of the port environment — enabling predictive maintenance for cargo handling equipment, real-time berth occupancy visualization, and active reduction of vessel turnaround time. This marks India's clearest smart-port AI integration milestone to date.
Organized by operational function — the current visible layer of India's maritime AI startup ecosystem.
In many shipboard fire incidents, critical minutes are lost before smoke reaches a detector. Dtyle.AI is positioning its onboard edge-AI platform to close that gap — but in shipping, the technology is only the beginning of the argument.
In many shipboard fire incidents, the critical minutes are lost before smoke physically reaches a detector — or before a crew member on rounds notices anything unusual. Machinery spaces, cable trunks, reefer decks, and restricted-access areas cannot be continuously monitored by human eyes alone. Night operations, high-workload conditions, and the geometry of large vessels create visibility gaps that conventional fire detection systems do not fully address.
That operational gap is where Dtyle.AI is positioning its onboard visual AI platform.
The company — incubated at IIT Madras — develops edge-based AI systems that analyse live camera feeds continuously for early indicators of smoke, fire, anomalies, and restricted-area breaches, before situations escalate into emergencies.
Traditional fire protection systems remain essential to SOLAS compliance. But most are reactive by design: smoke detectors activate only after combustion products physically reach the sensor. Dtyle.AI aims to move the detection timeline earlier, using onboard edge-AI processing that monitors visual patterns associated with smoke development, fire, or unusual activity.
According to the company, detection occurs locally onboard rather than relying on cloud connectivity — an operationally important distinction for vessels in low-bandwidth offshore environments.
The platform is designed for retrofit integration with existing IP-camera infrastructure, reducing installation complexity and minimising downtime. Dtyle.AI reports a pilot deployment onboard the Indian oil tanker Vamsee II, where real-time monitoring capability was achieved with retrofit installation completed within approximately two days.
For marine engineers, the important framing is that the system is positioned as an additional early-warning layer — not a replacement for statutory fire detection and firefighting arrangements, crew rounds, or permit-to-work disciplines.
Shipping earns its trust in systems slowly, and for good reason. Questions around false alarm rates, steam and aerosol interference, camera lens fouling, behaviour under blackout and power-restoration sequences, and cyber isolation from OT networks will ultimately determine whether visual AI becomes a trusted onboard tool — or simply another alarm source competing for attention.
The merchant fleet does not need less engineering judgement. It needs earlier visibility — before small abnormalities become major casualties.
As AI systems move onboard, marine engineers may increasingly find themselves responsible not only for machinery reliability, but for alert validation, camera system maintenance, cyber discipline, and the integration of AI-generated alerts into drills and emergency-response procedures.
Classification societies are shifting from periodic survey visits toward continuous data-based assurance, remote verification, and AI-assisted compliance monitoring.
| Society | 2026 Development | Source |
|---|---|---|
| DNV | Smart Shipping COE Phase Two (April 2026). RuleAgent AI regulatory navigation tool launched March 2026. Remote surveys and digital class notation. | dnv.com |
| Lloyd's Register | Live Mediterranean assessment of Orca AI situational awareness platform completed April 2026. Active in autonomous vessel assurance frameworks. | lr.org |
| Bureau Veritas | Continuing integration of remote survey capabilities and digital twin validation for operational vessels. | bureauveritas.com |
| ClassNK | Issued AUTO-Nav2(All) MASS notation to GENBU under MEGURI2040. Active in Japan's autonomous vessel regulatory development. | classnk.or.jp |
| IRS | India's national classification society and flag-state authority for Indian-flagged vessels. Central to any AI system onboard Indian-flagged ships. | irsofl.com |
Port AI adoption is maturing faster than onboard AI — with clearer commercial justification at every step.
AI systems may identify patterns. They do not replace engineering accountability.
The machinery did not become less important when computerized alarm monitoring appeared. The role of the engineer changed — more structured, more data-aware, but no less dependent on mechanical understanding, systematic thinking, and calm diagnosis under pressure.
AI is arriving in the same pattern.
It does not remove the requirement for an engineer who understands why a pressure trend looks the way it does. It does not replace the ability to distinguish between an instrument fault and a real process deviation. It does not remove the responsibility for decisions made in the engine room.
What it may change is where the engineer's attention is directed. Pattern recognition may become partially automated. Trend logging may be continuous rather than manual. Alarm correlation may be assisted. Voyage optimization decisions may arrive pre-calculated.
The verification — the discipline of checking, confirming, and understanding before acting — remains entirely human.
The marine engineer who learns to read data without losing practical judgement may become one of the most difficult professionals in shipping to replace.