Flight data monitoring (FDM) is an essential part of safety systems within flight operations. A wide range of aircraft parameters are recorded during flight, such as aircraft speed, altitude, rate of descent, position, engine speed, fuel flow, electrical parameters and aircraft configuration. Once on the ground, this data is analysed off-line to assess the performance of the aircraft and the fleet. FDM software currently makes use of maximum thresholds, which are flagged whenever exceeded. The data would require further investigation to establish if procedures or safety has been compromised. However, the ever-increasing sensors installation aboard aircraft, makes flight data analysis laborious and complex. Current techniques not only miss the ability to establish trends and patterns prior to developing faults. Furthermore, once an anomaly is established, a thorough investigative effort is required to determine the contributory factors leading to the anomaly. The latter heavily relies on the human expert but can be subjective to the trained user.

While the University of Malta has already done a substantial amount of work on using Machine learning techniques to detect anomalies in aircraft data, project WAGE consolidates these efforts by introducing the investigative aspect to establish the origins of the anomaly. The project therefore aims to establish contributory factors such as weather, geographical, and engineering parameters. To achieve this, the project aims to access freely available historical database of weather patterns and 3D geographical maps and merge them within a flight data analysis software. The project also aims to identify and cluster parameters which would highlight engineering concerns (such as requirement for early maintenance). This offers valuable insight for improved safety and flight operations. When applied within an environment of airline operations, it provides an acquired benefit of hindsight which can be referred to, to optimise future flight trajectories (for example when hazardous weather patterns are known and expected), and improve maintenance schedules with the minimum impact on the fleet operations.

Project WAGE adopts a multidisciplinary approach whereby research in AI is integrated into a setting of aircraft performance, gas turbine performance and airline operations. The project offers an innovative, step improvement over the current state of the art flight data analysis software. This may result in added value in improved aviation safety and airline economic advantages.