Gaining Actionable Insights Through Neurocomputational Trajectory Segmentation and Clustering

The key innovation of this project is the development of a NEurocomputational Trajectory Segmentation and Clustering (NETS) tool that will apply segmentation, explainable clustering, and unsupervised machine learning algorithms to gain actionable insights from large volumes of aircraft trajectory data. National Airspace System (NAS) trajectory data has all the characteristics of ldquo;Big Datardquo; such as volume, velocity, variety, and veracity and is widely available through services such as the System Wide Information and Management System (SWIM). Increased demand on the NAS and greater availability of data requires new tools and techniques to be developed to take full advantage of all available trajectory data. For this effort, Intelligent Automation, Inc. will develop the NETS tool to mine large volumes of trajectory data in order to gain actionable insights with the goals of improving aviation safety and efficiency, identifying anomalous and emergent behavior, and studying the impact of new entrants such as space vehicles, unmanned aircraft systems, and urban air mobility vehicles. Our NETS solution will apply state-of-the-art neurocomputational algorithms to partition a trajectory into meaningful segments and then group similar segments into clusters, thus enabling the automatic discovery of common, anomalous, or emergent movement patterns. The segmentation process ensures that meaningful trajectory segments are not missed, which could occur if a trajectory is considered as a whole. The NETS approach enables trajectories to be segmented and clustered in an unsupervised manner. Labels can then be assigned by a domain expert to each cluster to provide a classification. An explanation or rationale for why a trajectory segment was placed into a particular cluster will be provided by the NETS tool in order to facilitate the labeling process.