Schiphol: seeing the aircraft turn before the delay arrives
Camera observations and passenger-flow forecasts offer Schiphol two different ways to anticipate pressure. Autonomous baggage transport is being trialled, but a successful controlled route is not the same as operating across a busy apron.
Amsterdam Airport Schiphol apron with the control towers and a KLM Boeing 737, October 2011. Photograph: Bin im Garten / CC BY-SA 3.0 (cropped to fit). Shows the airport, not an AI implementation.
The picture at Amsterdam Schiphol
Where the clock starts running
An aircraft can be waiting for a final bag or a refuelling truck long before a delay appears on the departure board. Deep Turnaround was developed at Schiphol to make those intervals visible. Its camera-based system identifies ground-handling events and flags predicted late departures, giving teams information while there may still be time to respond.
The terminal poses another problem. A queue estimate is useful when it helps security teams deploy lanes and when passengers can plan their route through the airport. Schiphol has also tested a self-driving baggage tractor, although that remains a supervised trial rather than an airport-wide autonomous service.
These are separate operating problems with different measures of success. The sources establish working systems and trials, but they do not publish a Schiphol-wide before-and-after improvement attributable to the technologies.
Selected implementations
Two established uses, and a baggage experiment
01 / Deployed
A turnaround visible from the stand camera
The operational problem
The gaps between on-block and off-block times are operationally important, but often hard to reconstruct from ordinary flight milestones.
What the sources report
Schiphol's Aviation Solutions unit says Deep Turnaround identifies more than 70 events across 30 processes from camera images and can flag a possible delay around 30 minutes before target off-block time. The supplier offers the software commercially to other airports.
Why it matters
A timely warning might let handlers address a missed task before the departure is affected. Evidence from installations elsewhere cannot establish the resulting delay reduction at Schiphol.
What remains unproven
The number of Schiphol stands currently covered and Schiphol-specific before-and-after delay results have not been published.
Predicting a queue while there is time to staff it
The operational problem
Passenger arrivals at security do not follow a smooth daily average. Staffing decisions need a picture of both the present queue and what is likely to happen next.
What the sources report
Veovo's Schiphol case study describes live queue measurement using 3D and Bluetooth sensing alongside predicted waiting times displayed at processing points. A 2023 industry report says Schiphol was among customers migrated to Veovo's cloud platform.
Why it matters
Forecasts can help resource planning and passenger information. They should be judged against forecast error, queue waiting time and staffing decisions rather than the amount of data collected.
What remains unproven
Current queue-prediction accuracy and a measured operational benefit at Schiphol are not established by the reviewed material.
Evidence supports a reported claim; it does not automatically establish a measured operating benefit.
03 / Pilot
Baggage moves without a driver — but not without supervision
The operational problem
Transfer bags can arrive well before their onward flight, creating extra movements between aircraft and temporary storage.
What the sources report
Schiphol and KLM began testing Aurrigo's all-electric Auto-DollyTug in 2024. Aurrigo reported a supervised route trial in 2025, with LiDAR, cameras and an operator on board who could intervene. The partners proposed progressively busier routes and possible aircraft-stand tests.
Why it matters
The operational interest lies in whether repeatable baggage movements can be handled safely with less manual driving. A staged trial is appropriate evidence of experimentation, not of fully autonomous fleet operation.
What remains unproven
Published evidence reviewed for this profile does not establish the later stand-test outcome, ongoing deployment scale or productivity improvement.
Evidence supports a reported claim; it does not automatically establish a measured operating benefit.
The wider register
What else is happening at Amsterdam Schiphol?
Other Schiphol records include security CT equipment. Such scanners are important airport infrastructure, but the sources reviewed do not establish that Schiphol's CT screening is an AI deployment. Cancelled projects remain excluded from the public register.
CT scanners at all Schiphol departure and transfer security filters · Deployed · full scale
Every departure and transfer security filter at Schiphol has used CT scanners since May 2021, when a new filter opened on the mezzanine of Departure Hall 1, the airport announced on 6 May 2021. Schiphol described itself as the first major airport to make the full switch, so electronics and liquids can stay in bags.
Lane supplier Scarabee, part of Daifuku Airport Netherlands, says 85 of its Smart Security Lanes have been installed across Schiphol's checkpoints since 2015, and that all its Smart Security Lanes carry CT scanners. The 85 is the lane vendor's count, and Schiphol's announcement does not name the scanner manufacturer.
Neither Schiphol nor Scarabee calls the screening AI, and no detection software has been named. Throughput and alarm-rate data have not been published.
Deep Turnaround and queue prediction share a practical purpose: give operational staff useful notice. The evidence establishes systems designed to do that, not a measured airport-wide reduction in delays or waiting time.
The baggage vehicle is different: its present value is in what a closely supervised trial can establish about safe routes and handovers. That distinction between operating tool and experiment is worth preserving.
Live data from the Aerostack register. Status reflects published evidence only. See Our approach for how records are graded.
Aerostack register · 4 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.
