Skip to content
Artificial intelligence. Clearly. RSS feed
AI NEWS AI

News. Evidence. Perspective.

BusinessAnalysis

How a faster port can give a ship time to slow down

Maersk’s AI story runs through cranes, container yards and truck schedules. Its most useful lesson is how a better plan at one handover can change the economics of the next.

A ship that finishes its port call sooner may be able to sail more slowly to its next destination. That is the connection Maersk draws between terminal productivity and fuel efficiency: saving time alongside the quay creates room to reduce speed at sea. The interesting unit of improvement is the journey across both operations.[4]

In an interview published by CIO/ISMG on 16 October, Maersk technology executive Resham Sahi described AI returns in physical terms: vessel sailing time, terminal turnaround and fuel use. She discussed digital twins that bring live operational data together with booking information and terminal systems to improve such decisions as rostering and movements around a container yard. Her account puts a practical question at the centre of the investment: can a prediction become a workable plan?[1]

Thirteen systems between a container and its truck

Maersk’s published account of its Southern California drayage operation shows what that question involves. Drayage is the short truck journey connecting a port with another logistics facility. Frontline planners were using 13 systems, including terminal, transport-management, chassis-provider and yard systems. A container’s status existed in pieces; constructing the next move meant joining those pieces by hand.[2]

The deployed digital-twin project draws information into a common platform using APIs where available and robotic process automation for some legacy systems. It combines vehicle and shipment information with traffic inputs, then optimises work subject to operational constraints. An electric truck adds another set: remaining range, charging and the downtime charging entails. The plan must choose drivers and vehicles as well as the order of collections and deliveries.[2]

Consider an illustrative handover. A crane unloads a container earlier than expected, but the allocated truck is still collecting another load. Speeding up the crane has moved the waiting time onto the quay. A useful planner must ask whether another truck can take the job, whether its next appointment still works and whether moving this container delays a more urgent one. The value lies in a feasible sequence of actions.

A forecast has to survive the handover

Maersk was describing terminal digital-twin rollouts in January 2023. Container volumes, vessel schedules, sensors and terminal data fed simulations of near-term operations, helping anticipate which resources would be needed where. The company identified uncertain task durations as a source of inefficient buffers: a crane can be available, yet available at the wrong time. It also said a connected twin of its entire business was years away. That older account helps distinguish an established programme of individual applications from a claim that every link has already been unified.[3]

The distinction matters commercially. Forecasting a vessel’s arrival is one task. Reserving a berth, arranging the right crane capacity and making a workable landside plan are connected decisions. A local optimisation can improve one team’s statistic while leaving the customer’s shipment no earlier. The stronger business case measures what happens across the handovers.

Where the saved time goes

Maersk’s 2025 annual report says APM Terminals hubs increased productivity, measured through port moves per hour, by 14% over two years, with port stays 15–20% shorter. These company-reported gains cover a wider equipment, capacity, technology and operational programme, without isolating AI’s effect. The port-stay reduction is also not a fuel-saving percentage.[4]

The same report names Port Mirror, a digital twin used to simulate terminal operations and anticipate congestion, and Berth Planner, which helps organise vessel line-ups and berth use. It connects better port-call performance with more sailing time at lower speeds and less variation in speed.[4]

The instructive part for an AI buyer is the connection between operations. A berth-planning tool matters when time saved at the port becomes time the ship can actually use.

Follow the evidence

Sources & further reading

  1. Measuring AI ROI in Fuel, Not Seats CIO / ISMG · 16 Oct 2026 · Accessed 18 Oct 2026
  2. Digital Twin Maersk Innovation Center · Accessed 18 Oct 2026
  3. Annual Report 2025 A.P. Moller - Maersk · 5 Feb 2026 · Accessed 18 Oct 2026