Ürümqi: driverless tractors become part of the freight operation
Ürümqi Tianshan has one of the clearer reported cases of autonomous ground vehicles moving substantial cargo volumes, alongside a new terminal with AI-assisted operational and security systems. Most performance numbers come from suppliers or local reporting and require careful attribution.
Terminal of Ürümqi Tianshan International Airport. Photograph: HaziiDozen / CC BY-SA 4.0 (cropped to fit). Shows the airport, not an AI implementation.
The picture at Ürümqi Tianshan
Automation operating through a cold-weather cargo hub
Airport autonomy is often judged by whether a vehicle can navigate a test route. Ürümqi's more consequential claim concerns a fleet repeatedly carrying cargo and baggage in a working airport, including severe winter conditions.
UISEE says its autonomous-tractor fleet expanded from 40 vehicles in 2025 to 52 by August 2026, handling a large share of cargo and mail movements. The north terminal, opened in April 2025, introduced a different set of technologies including intelligent load planning, security-image assistance and digital-twin representations.
Those accounts suggest a broad operating programme, but the fleet percentages are vendor figures. The reported terminal improvements cannot automatically be attributed to one AI tool.
Selected implementations
From autonomous towing to terminal decisions
01 / Deployed
The driverless tractor that moves real freight
The operational problem
Cargo and baggage transfer requires repetitive movements across an apron, with safety and reliability maintained in difficult weather.
What the sources report
UISEE reported 40 autonomous tractors in 2025 and 52 in 2026, claiming the later fleet handled more than 85% of cargo and mail and over 75% of baggage transfer tasks. Other Chinese reporting describes a larger mixed autonomous fleet.
Why it matters
The reported share of actual transfer work, if verified, would make this an important operational autonomy case.
What remains unproven
Independent measures of the percentages, supervisor interventions, incidents and comparative operating costs are unavailable in the attached evidence.
Evidence supports a reported claim; it does not automatically establish a measured operating benefit.
02 / Deployed · limited scope
The new terminal's operational nervous system
The operational problem
Stand allocation, flight delays and security rechecks need coordination across systems that often work on different information.
What the sources report
CAAC News described TravelSky's suite introduced with the new north terminal in April 2025, including AI-supported flight-delay management, security quality checks and automatic load planning.
Why it matters
The system is significant because it joins several operational decisions, although each function has a different maturity and human-control requirement.
What remains unproven
The current sources do not independently isolate AI's impact on stand allocation or screening reliability.
Evidence supports a reported claim; it does not automatically establish a measured operating benefit.
The wider register
What else is happening at Ürümqi Tianshan?
Face-recognition security verification gates and tracking of late passengers are recorded separately, including historic processing-time figures. They should not be presented as evidence of the tractor fleet's safety or benefits.
Face-recognition self-service security verification gates and late-passenger tracking at Ürümqi · Deployed · limited scope
Ürümqi airport put 37 face-recognition self-service security verification gates into use in June 2020, cutting verification from 25 seconds to under 8 seconds per passenger, CAAC News reported in August 2022. The airport also uses camera face capture to locate late passengers for boarding calls.
Its north terminal, opened on 17 April 2025, added two-way face-recognition boarding gates.
An operating share is a stronger claim than a robot count
Ürümqi has enough reported operational use to merit serious attention. The next evidential step is to corroborate how much freight the vehicles actually move.
Fleet size, claimed percentages and terminal-system descriptions should remain attached to their respective sources, not turned into one airport-wide AI performance figure.
Live data from the Aerostack register. Status reflects published evidence only. See Our approach for how records are graded.
Aerostack register · 3 tracked deployments
Aerostack tracks AI deployments across airports and airlines, with the published evidence behind each record. Photography is real aviation imagery. Record thumbnails may show the named airport or airline; other photographs do not.
