The tutorial introduces several Air Traffic Management (ATM) initiatives envisioned by the Federal Aviation Administration (FAA) for a bold future airspace that combines conventional traffic and new entrants (e.g., drones) without sacrificing safety. In this framework, we demonstrate the use of state-of-the-art AI/ML modeling and prediction tools that will enable efficient and safe traffic flow in the U.S. National Airspace System (NAS). In particular, Natural Language Processing (NLP) tools can help extract data (e.g., airspace constraints) that are currently contained in legacy text and audio format and convert them into digital information. The digitized information can be ingested by route planning, arrival scheduling and other decision support tools both on the ground and in the flight deck.
We also show how historical data (track, weather & events) can be preprocessed, cleaned and utilized to create accurate models to predict flight trajectories and events of interest (e.g., Traffic Management Initiatives). We show several application areas within ATM that benefit from AI/ML including trajectory prediction, airport runway configuration management, automatic speech to text (of FAA command center webinars) and digitization of Letters of Agreement. The overarching goal of the work is to accelerate the integration of package delivery drones, air taxis and autonomous cargo aircraft into the NAS without impacting the safety and efficacy of current manned operations. With the correct application of modern AI/ML tools and availability of abundant data (both structured and unstructured), it is possible to build accurate models to do both prediction (e.g., estimated time of arrival) and provide recommendations (e.g., which runway configuration should be used). This tutorial is tailored to educate students, researchers and practitioners on what NASA is doing in this problem space and how AI/ML can help in solving some challenging problems in ATM. We will showcase several AI/ML methods used e.g., random forest, XGBoost, model free RL and NER by showing code snippets, input data samples and example output/results that validate our research.
