Job Description

Project Information:

This project seeks to investigate and demonstrate methods to enable rapid selection of days for scenario generation in the development and evaluation of air traffic management (ATM) concepts and technologies (C&T). The proposed capability will enable the rapid generation of highly operationally relevant scenarios for use in the development and evaluation of technology demonstrators such as NASA Airspace Technology Demonstrator (ATD)-2 and ATD-3, Unmanned Aerial Vehicle Traffic Management (UTM), as well as new operational concepts such as Integrated Demand Management (IDM) and Trajectory Based Operations (TBO).

The proposed innovation seeks to augment the scenario generation capability with methods and tools for selecting traffic, winds, and weather based on the needs of the experiment allowing for highly operationally relevant scenarios. These methods and tools would actively categorize incoming and historical data using advanced machine-learning algorithms, allowing fast access to NAS streaming and legacy data in a big-data warehouse through queries generated via a simple user interface for specifying desired characteristics.

The motivation for this innovation is that over the past several years, because of the decreasing cost of storage, large volumes of aviation related data have been collected by several organizations including NASA and the Federal Aviation Administration (FAA). While the utility of archiving these data is clear, their real use by researchers and projects has been limited. NASA has recently invested in cleaning up and improving the consistency of the archived data. This has made it easier to process the archived data further. Users can retrieve archived data from the Data-Warehouse via the web page, provided they know the particular day’s data of interest. However, because days of interest are difficult to identify, researchers typically base their experimental evaluations on only a few days of data. The evaluation of a concept or technology’s system-wide impacts in terms of cost and benefits with one or two days of data, or with several days of data with similar traffic conditions, is of limited utility. Such evaluations should instead be conducted with a set of days with distinct characteristics. The proposed innovation seeks to overcome this limitation by providing an automated way for identifying days of interest based on high-level characteristics specified by the user.