The picture at Keflavík International
When connecting passengers arrive in waves
Keflavík's transfer passengers do not arrive evenly throughout the day. Short connection windows make it important to know how many travellers are approaching a processing point before the queue appears.
Veovo says Keflavík adopted its Passenger Forecasting product in 2020 and later deployed an AI-powered Queue Balancer. The supplier describes a practical link between demand estimates and flow decisions.
The published record establishes named systems and reported deployment. It does not supply an independently measured improvement in missed connections or passenger waiting time.
Selected implementations
The airport forecast and its operational use
01 / Deployed
Predicting the transfer wave
The operational problem
Resource plans can fall behind when several incoming flights deliver passengers into the same constrained connection period.
What the sources report
Veovo's July 2020 announcement describes Keflavík as the first deployment of its Passenger Forecasting product. Its June 2023 case study describes the airport implementing an AI-powered Queue Balancer for transfer flows.
Why it matters
The useful measure is whether forecast information changes queue management early enough to protect connections.
What remains unprovenVeovo is the source for both deployments; independent predictive accuracy and observed passenger benefits are not established.
Sources and verification (2)
Veovo case study · Evidence grade B
Veovo · Evidence grade B
Evidence supports a reported claim; it does not automatically establish a measured operating benefit.
The wider register
What else is happening at Keflavík International?
The current register groups Veovo forecasting and queue balancing in one related programme. They should not be counted as two independent airport-wide AI installations.
No additional publicly listed deployment records.
A forecast must be judged against a connection
Keflavík's operational question is unusually clear: how to move transferring passengers through the airport when flights arrive in clusters.
Evidence from the airport on queue variability and missed connections would materially strengthen this profile.