Description: This SBIR system is a machine learning system that uses a very large database of airside and landside data to predict pushback and takeoff times of aircraft at a given airport. Airside data sources describe the state of the system after TSA security screening is complete, and includes information about the crew and passengers arriving at the departure gate, turnaround and pushback preparation, ramp and taxiway movement, and aircraft arrival to and departure from the gates. Landside data sources describe the state of the airport prior to TSA screening, including TSA queue line delays, passenger movement through the airport via cameras, parking availability, road transit delays, congestion, and accidents, and weather conditions. These data are used to classify the current day data using cluster analysis, and take off time and pushback time predictions are made based on the cluster analysis results.

Benefits: This work is fundamental to Air Traffic Management (ATM), and will naturally fit into ATM research being performed at NASA Ames and NASA Langley. It will likely be used in ATM research efforts, Trajectory-Based Operations (TBO) research, and in the SMART NAS system under development at NASA.

Based on a partnership with Metron Aviation, plans are to include this SBIR software into products and services that Metron Aviation sells in the National Airspace System (NAS) as well as throughout the world.