Validation Framework for Automated Aircraft Load Control Operations
Abstract
This paper presents a validation framework for automated aircraft mass and balance operations, combining statistical process control with integrated risk scoring and addressing the critical gap between traditional centralised load control and emerging automated operations. A dual-validation system integrates univariate control limits, a weighted multivariate T2-type statistic, and composite scoring with operational risk assessment; risk scoring quantifies centre of gravity proximity and maximum mass margins. The framework was validated using 48 flights operated by Airbus A350-900/1000 aircraft. The framework was evaluated using a two-phase design—flights 1–30 for calibration (Phase I) and flights 31–48 for prospective evaluation (Phase II). Across all 48 flights it yields an 81.25% automation-eligibility rate (39 GREEN), with 4.17% YELLOW (non-blocking supervisor notification within a defined service window) and 14.58% RED (mandatory review). RED cases result from cargo offloads, dangerous goods changes, or significant baggage/ZFW variations, while cargo operations show exceptional stability (the exclusively cargo-loaded Compartment 2 has σ = 9 kg, versus 219 kg for total cargo). The framework enables transition from centralised load controller roles to exception-based supervision, with risk scoring outputs driving targeted, proportional safety measures. To the best of the authors’ knowledge, it provides the first systematic method for determining when automated load control can be trusted without human intervention—a critical requirement as the industry transitions towards full automation.