Air Traffic Management Simulation Data Visualization and Processing Tool

Abstract

Large-scale distributed simulations in which pilot participants fly simulators through multiple airspace sectors managed by controller participants create a rich operational environment for investigating new air traffic management concepts. This paper describes a Java™-based tool that aids in integrating, visualizing, and transforming data collected from large-scale human-in-the-loop air traffic management simulations. I. Introduction UTURE Air Traffic Management (ATM) concepts commonly leverage flight deck and ground-based automation and new interface tools to improve efficiency. To evaluate the feasibility and benefits of proposed concepts, researchers need ways to assess changes in practitioner roles and responsibilities, in addition to traditional system performance, human performance, and acceptability metrics. One way to investigate air-ground integration issues is to use large-scale simulations in which pilot participants fly simulators through multiple airspace sectors managed by controller participants, creating a rich ATM environment. This paper describes a Java™-based tool called DProc that aids in visualizing, integrating and transforming data collected from large-scale ATM simulations. DProc was implemented to process data from simulations in the Airspace Operations Laboratory (AOL) at NASA Ames Research Center. Controller stations, piloted and pseudo-piloted aircraft simulations—as well as simulation manager and ‘host computer’ components—all produce data describing particular aspects of an overall simulation. A given controller station, for example, logs actions performed by that air traffic controller subject, while the host computer provides a repository for aircraft state and trajectory information and logs descriptive data. Researchers may be interested in tracing event sequences, information flows, and operational contexts associated with certain outcomes of interest. This may entail identifying what other human subjects or automation agents are doing when one performs an action, and measuring relationships such as the time between various events or actions. Analysts can, of course, use traditional tools like spreadsheets to examine ATM simulation data and analyze interactions. However, the different types, formats, and sheer quantity of data can make this difficult. Moreover, data collected in a research environment with prototype automation tools may require integrity checking and other preprocessing steps prior to analysis. DProc addresses these issues by creating a database of merged simulation data from various sources. The DProc interface enables researchers to replay simulated traffic and visualize recorded events together with aircraft states. Visualization data may be color-coded according to traffic characteristics (e.g., aircraft weight class, equipage, engaged autoflight modes), or filtered to highlight events associated with a particular controller, aircraft, or class of aircraft. In addition to replaying data, DProc is also capable of producing plots of aircraft tracks or event locations. In addition to visualizations, DProc also produces batch output suitable for input to a spreadsheet or other analysis tool. As an example, the paper notes how DProc produces input files for data mining applications. Finally, the paper describes potential enhancements to DProc. For example, audio and video, as well as other subjective data collected during or after a simulation run might also be integrated and used to support analysis.