Visualizing Corridors in Terminal Airspace using Trajectory Clustering

Context: Advances in battery and automation technology have made routine air taxi and cargo transport in urban areas a business model that can be attained by emerging aviation innovators. The community vision and work to enable these novel operations is discussed using the term ‘Urban Air Mobility’ or UAM. Small, piloted, airspace vehicles that fly with a few passengers do operate in urban areas today, and these vehicles can be studied as an early proxy for this future UAM traffic.Aim: We seek to identify corridors already in daily operation and their properties.Method: We applied DBSCAN and HDBSCAN to Dallas Forth-Worth TRACON flight data to identify corridors in use, their density, and devised a method to annotate landing sites used in these corridors with site metadata.Results: While DBSCAN was unable to group similar trajectories, we we were able to successfully identify corridors using HDBSCAN, measure their density and annotate them.Conclusion: The applied method can successfully identify corridors in daily operation with additional metadata to help domain expert understand the intent of UAM corridors.