Worldwide Flight Services (WFS) has upgraded its machine-learning–driven forecasting platform to deliver sharper, more reliable predictions of air cargo volumes and strengthen operational planning across its global network.
The Performance Management Platform – Machine Learning Forecast (PMP MLF) now analyses a decade of operational data to predict cargo volumes by flight, truck, and day. Trained on more than 3 million air waybills and extensive historical movement records, the system factors in seasonality, holidays, and cargo types to provide warehouse-level insights that help optimise labour and resources in advance.
The tool currently generates daily forecasts for 9,842 flights and 6,216 truck movements each week, covering 75 warehouses in 13 countries. It breaks down tonnage, ULDs, and piece counts by transport mode, customer, and location, feeding this intelligence directly into station-level planning tools.
With the enhanced system, WFS can detect surges earlier and redeploy staff more efficiently across teams and sites, reducing SLA breaches and avoiding both overtime costs and idle time.
Phase two of the platform, rolled out this summer, introduced improved dashboards, deeper visual analytics, tighter links with workforce management and rostering systems, and customer-specific forecasts for joint peak planning.
WFS reports that the upgraded tool now delivers 92–98% accuracy, even during volatile demand periods—helping address one of the air cargo sector’s long-standing challenges: the industry’s reliance on manual estimates that often leave a 10–15% gap between required staffing and actual workload.
