A.P. Møller-Maersk, the Copenhagen-based container shipping giant with $2.3B fourth-quarter 2025 revenue and 100,000+ employees globally, deployed Star Connect, an edge-computing AI platform running directly on board vessels processing 2.5 billion IoT data points from onboard sensors, engine performance readings, weather feeds, and current data.
ML models run locally on each ship's server because satellite bandwidth at sea is too expensive and slow for cloud processing. The platform delivers up to 15% fuel savings on optimized routes, with Maersk spending $5-7B annually on bunker fuel. The Gemini alliance network planning engine targets 90% schedule reliability, nearly double the industry average.
The shift to edge AI on vessels addresses a fundamental constraint: shipping operates at sea where connectivity is expensive and intermittent. By running ML inference locally, Maersk can optimize fuel consumption and predict maintenance needs in real time without depending on cloud connectivity. AI-driven route optimization has achieved 9.2% reduction in fuel consumption while predictive maintenance has cut vessel downtime by 30%, contributing to over $300M in annual savings.
The deployment represents Maersk's strategic pivot from attempting to be the operating system of global shipping (TradeLens shutdown 2023) to deploying AI inward on its own operations. The AI-Powered Vessel Routing Platform developed with Microsoft Azure AI processes real-time data across the fleet, and the OneWireless connectivity platform supports both Star Connect and predictive maintenance systems.
Architecture spans on-vessel edge servers running ML inference locally, IoT sensors capturing 2.5B data points across onboard systems, the Microsoft Azure AI-powered Vessel Routing Platform, the Gemini alliance network planning engine, and the OneWireless connectivity platform supporting Star Connect and predictive maintenance.
Operational scale covers 450+ vessels in the Maersk fleet, processing 2.5 billion IoT data points, achieving up to 15% fuel savings on optimized routes and contributing to over $300M in annual savings. Maersk's $5-7B annual bunker fuel spend makes even single-digit efficiency gains highly material. The Gemini alliance targets 90% schedule reliability.
Looking ahead, Maersk plans integration with new dual-fuel methanol fleet, expansion of generative AI for complex planning, and conversion of internal AI tools into the customer-facing Maersk Trade & Tariff Studio. Maersk also announced 1,000 corporate layoffs in February 2026 linked to AI-driven automation targeting $180M in annual savings.