Global Technology Connection, Inc. and its partners propose to develop a flexible state-of-the-art capability focused on identification of trajectory anomalies, based on energy metrics, using a fusion of data from multiple sources (e.g., OpenSky Network data, Traffic Flow Management System Data, Automated Surface Observing System Data) in order to identify and isolate potential causal factors and/or anomalies, and help improve safety of operations in terminal areas.
Future aviation systems such as the NextGen, and Single European Sky ATM Research (SESAR) are moving towards utilize 4D trajectory management concept in order to improve efficiency, reliability, sustainability and cost-effectiveness of aircraft operations. This will require aircraft to follow an assigned 4D-trajectory (time-constrained trajectory) with high precision. By detecting trajectory anomalies, safety-critical risks include flight outside of approved airspace, unsafe proximity to people/property, loss of command, control, power, loss or degraded GPS, and engine failure can be identified in real time to be used for prognosis and resolution of system-wide threats.
Despite numerous implementations of anomaly detection using flight data, there are limited frameworks that fused data from multiple sources (weather information, system level metrics related to congestion, traffic, etc.), which is what we are proposing by performing the following technical tasks: Identification of data sources, Data fusion, Deep autoencoder model development, Post-processing.
In phase I, we develop a prototype for proof-of-concept. In Phase II, we add other data sources and develop a commercial prototype for NASA applications and design a verification & validation process to meet the NASA requirements. In Phase III, we develop the commercial version of the software package for NASA and Non-NASA applications, and continue commercialization of the intended product.