Artificial intelligence and robotics are moving from pilot projects into day-to-day logistics operations, with companies applying the technologies to transport planning, warehouse automation, maritime shipping and air freight, according to a report from delivery services company Parcelhero. The report says autonomous mobile robots, predictive systems and AI-based planning tools are becoming embedded across supply chain operations.
The report, titled Putting the AI into Supply ChAIns, says the proportion of UK transport and storage businesses using AI increased from 16.1% to 27.1% during the first three months of 2026. More than 20% of the remaining companies were planning to adopt the technology during the second quarter, it said.
Parcelhero cited estimates showing that about 40% of supply chain organisations worldwide are investing in generative AI. It also referenced Deloitte’s 2026 Retail Industry Global Outlook, which found that 41% of retailers planned to deploy AI within 12 months to improve supply chain visibility.
The report identified route optimisation as one of the main areas of adoption. It said AI-based systems had reduced logistics costs by 10% and improved on-time delivery rates by 15% in some applications. Such systems can process traffic conditions, weather information and customer preferences to reorder routes containing more than 100 delivery stops.
AI is also being used to forecast shipment volumes and manage the final stage of deliveries. Parcelhero said some systems could predict volumes for individual facilities with accuracy of up to 95%, allowing operators to position inventory in advance of expected demand. The report estimated that the use of AI in last-mile delivery had increased by 39% over the previous year.
Warehouses are increasingly using autonomous mobile robots alongside predictive maintenance software and energy-management systems. Citing McKinsey, the report said 56% of businesses had integrated AI into at least one operational function. AI-enabled warehouse systems can adjust robot routes in response to changing conditions, identify potential equipment failures and manage electricity consumption.
Adoption is also increasing in maritime transport. Parcelhero valued the maritime AI market at £4.13 billion in 2024 and said it was projected to grow at a compound annual rate of 23% over the following five years. The number of organisations adopting maritime AI technologies rose from 276 in 2024 to 420 in 2025, according to the report.
The report cited the Yara Birkeland, an electric autonomous container vessel operating in Norway, as an example of automation in commercial shipping. It said the vessel had completed more than 250 voyages and was expected to replace about 40,000 diesel truck journeys a year on its route.
In air freight, AI is being applied to rate comparisons, export classifications, customs documentation and estimated arrival times. Automating these processes can reduce the time required to complete administrative and planning tasks while helping operators identify potential disruptions earlier.
Parcelhero said adoption remained constrained by legacy technology, batch-processed data and manual workflows that cannot support the real-time information required by many AI systems. Data quality, digital infrastructure and system integration would therefore be central to further deployment.
Only 16% of logistics and supply chain organisations surveyed were unlikely to adopt AI within five years, according to the report. However, it said differences in technical readiness could affect how quickly companies introduce the technology.
The report also examined the effect of AI adoption on employment, citing Office for National Statistics data showing that 31% of transport and storage businesses using AI had reported no change in staffing levels. The number reporting confirmed job reductions was too small to be statistically recorded, it said.
David Jinks, Parcelhero’s head of consumer research and author of the report, said companies were increasingly embedding AI, automation and robotics into operations rather than limiting their use to trials. He identified workforce training, data quality and digital infrastructure as the main requirements for companies seeking to expand their use of the technologies.
