Elida C. Smith

Applying automatic speech recognition technology to Air Traffic Management

Controller-pilot voice communications are a critical component of the Air Traffic Control (ATC) system, but outside of the human listening and responding that occurs with each transmission, they are an underutilized source of information for automation systems in the ATC domain. Automatic speech recognition is a continuously improving technology that can be used to tap into this information source for potential system benefits in a variety of ATC applications, such as monitoring live operations for safety benefit, conducting analysis on large quantities of recorded controller-pilot speech, or enabling automated simulation pilots to facilitate training and Human-in-the-Loop (HITL) simulation experiments. This paper describes how automatic speech recognition can be used in the ATC domain, the characteristics of the automatic speech recognition process and the ATC domain that make the problem unique, and the engineering process for effectively applying automatic speech recognition to ATC systems.

Eye-Tracking Analysis of Near-Term Terminal Automation for Arrival Coordination

Sponsored by the Federal Aviation Administration (FAA), The MITRE Corporation’s Center for Advanced Aviation System Development (CAASD) developed the Relative Position Indicator (RPI) concept. RPI is an automation concept to aid air traffic controllers in coordinating arrival traffic, reducing the need to vector for spacing during merging operations and, thus, retaining the benefits of Area Navigation (RNAV) and Required Navigation Performance (RNP) procedures. Validation activities involving an RPI research prototype have shown benefits in achieving more efficient merging and spacing of aircraft. Since the RPI concept makes additional information available on the situation display that can be referenced by the controller, an understanding of this information presence upon controller scan behavior was desired. As part of a Human-In-The-Loop (HITL) simulation of Denver operations conducted by CAASD, a comparison of two scenarios was made to evaluate the change in controller attentional allocation , as measured by an eye-tracking capability, when RPI is introduced. The simulated traffic consisted of predominantly RNAV operations and was managed with and without RPI automation. Participants were air traffic control specialists at Denver Terminal Radar Approach Control (TRACON) facility. Given the design of the RPI tool, two behavior changes were anticipated; 1) an increase of time spent scanning the primary flow of a merging flow geometry and 2) an increase of time spent scanning farther ahead or in advance of the merge point location. The eye-tracking analysis provides a preliminary indication that RPI does affect air traffic controller visual scanning patterns. Although the magnitude of change does not seem concerning , additional evaluation should be pursued to understand the impact of this change.

Analysis of controller-pilot communication performance in Area Navigation (RNAV) and conventional arrival operations

Reductions of controller-pilot communications and workload have been shown as a result of area navigation (RNAV) procedure implementation in the terminal environment, yet the impact on controller-pilot communication performance aspects, such as the lag between communication issuance and response, the duration of communications, and the time spent communicating, is unknown. It was hypothesized that the combination of reduced communications and workload (enabled by RNAV implementation) could result in reduced aural vigilance (which would manifest in the form of an increased response lag). It was also hypothesized that little change in the amount of time spent issuing and responding to communications would be shown due to the adjustment of human behavior to a lower level of workload experienced with RNAV operations (which would manifest in the form of decreased speech rate and/or increased use of chat (e.g., salutations)). To investigate these effects, controller-pilot voice communication field data reflecting busy arrival traffic was evaluated in conventional operations and RNAV operations. Pilot and controller communications were broken into three components; (1) the initial communication (made by either the controller or the pilot), (2) the response communication to the initial communication (made by either the controller or the pilot), and (3) the time lag in between the initial communication and the response communication. Results showed that controller and pilot communication duration, the time between the initial and response communications, and total time-on-frequency were all relatively unchanged across conventional and RNAV arrival operations.

Assessment of controller situation awareness in future terminal RNAV operations

The MITRE Corporations Center for Advanced Aviation Systems Development (CAASD) was tasked by the Federal Aviation Administration (FAA) with defining and validating the performance-based air traffic management (ATM) concept to address the increasing need for improved capacity, efficiency, and productivity in the National Airspace System (NAS). A key enabler of this concept is the continued implementation and greater utilization of performance-based navigation provided by area navigation (RNAV) and required navigation performance (RNP) today and through the future. Implementation of new procedures and features, such as vertical profiles, are expected to reduce the active controlling task that is fundamental to air traffic control (ATC) operations and instead, increase monitoring of operations. This change is intended to leverage flight deck automation and reduce pilot and controller workload; however, it is critical to fully understand the human factors implications of making this shift, especially as traffic operations continue to grow. A human-in-the-loop (HITL) simulation was performed to evaluate changes in situation awareness for the feeder arrival controller position in a terminal radar approach control (TRACON) environment with conventional arrival operations and with future RNAV arrival operations when moderate and high levels of traffic were managed. An assessment of controller situation awareness was made based upon results from the situation awareness global assessment technique (SAGAT) and measurement of controller detection of pre-planned aircraft deviations from their assigned clearance. Workload was also measured using the NASA-task load index (TLX). Findings suggest that a change of controller situation awareness does occur as a result of the increased traffic levels managed. Assessing and mitigating this issue is under study by MITRE/CAASD, in particular through display and alerting automation for radar controllers.

Ground and flight deck alternatives for terminal merging, sequencing, and spacing for arrivals

Aircraft approach the terminal area on different arrival paths, merge and ultimately arrive at capacity-constrained runways on the ground. Upon entering the terminal area, air traffic control manually controls the aircraft in order to merge and sequence arrivals on different paths, and establish longitudinal spacing between aircraft. These operations are termed merging, sequencing, and spacing operations. New technologies available in air traffic control facilities and in the flight deck create the possibility of improving on the manual operations used today. Several alternatives are currently in operational trials. Unclear is what alternative will prove to be most desirable for implementation. This paper seeks to enumerate some alternatives, and their characteristics, and establishes a framework for assessment.

Requirements elicitation through storyboarding NextGen operational improvements

The Federal Aviation Administration (FAA) has defined several operational improvements, based upon emerging technologies and aircraft equipage, which largely comprise air traffic operations expected at the end of the mid-term time frame, 2018. These operational improvements are intended to not only enable user and system-wide benefits, but also serve as the foundation for transforming the National Airspace System (NAS) into the Next Generation Air Transportation System (NextGen). To plan for and better understand operational improvement implementation issues, the MITRE Corporations Center for Advanced Aviation System Development (CAASD) has been formulating scenarios, derived from identified sets of related operational improvements, that provide a snapshot of future operations and detail key aspects. Development of the scenarios was largely based on the utilization of a visually motivated technique referred to as ‘storyboarding’. The nature of storyboarding allows for new concepts to take a ‘shape’ that can be commonly understood. This affords consensus building and promotes feedback acquisition in the early stages of the design phase, both which were prominent goals of this effort. These goals are accomplished because the scenarios developed provide a vehicle for conceptualizing interactions between the operational improvements described in the NextGen Implementation Plan on a more tangible level while also communicating their impact to various aviation and industry stakeholders. Scenarios can also facilitate cross-FAA and industry-FAA discussions for the purpose of identifying implementation gaps and, in turn, stimulate collaborative solutions which will serve to define implementation requirements. The compilation of these scenarios will help clarify the impact of operational improvements and their utilization. More importantly, the collaboration among key stakeholders as the by-product of this process will serve as the basis for successful partnerships in the FAAs progress towards NextGen.

Tower controller response to runway safety alerts

The contents of this material reflect the views of the author and/or the Director of the Center for Advanced Aviation System Development. Neither the Federal Aviation Administration nor the Department of Transportation makes any warranty or guarantee, or promise, expressed or implied, concerning the content or accuracy of the views expressed herein.

Use of near-term terminal automation capabilities for meeting an evolving operating environment

The Next Generation Air Transportation System (NextGen) will use technological updates and systems engineering methods to improve the efficiency, safety, and capacity of the National Airspace System (NAS). While some of these changes may be centered in the Air Route Traffic Control Center (ARTCC) or the Airport Traffic Control Tower (ATCT), it is important to evaluate the operational effects on Terminal Radar Approach Control (TRACON) Air Traffic Control (ATC) tasks so that TRACON automation investment planning will proceed in the most effective manner. Controller task changes and new operational challenges may be a result, with some addressed by using automation capabilities in the current baseline or planned near-term automation capabilities. In order to identify potential operational challenges, The MITRE Corporations Center for Advanced Aviation System Development (MITRE/CAASD) worked with TRACON ATC Subject Matter Experts (SMEs) to analyze planned NextGen improvements, and discuss how the future operating environment would differ from the current one [1]. Additionally, the group identified ways in which current and/or near-term automation capabilities could be used to address the identified challenges. Some findings from this effort are detailed in this paper.

Controller response times to surface safety alerts

To ensure that Air Traffic Control Tower-based surface safety systems are designed to appropriately take into account human response time, it is critical to understand controller responses in situations that are within the controller’s operational envelope (i.e., situations that the controller would be expected/trained to handle as part of the job) but that are expected to elicit response times longer than those that would be expected in more typical air traffic control situations. This paper describes the conduct and results of a Human-in-the-Loop (HITL) simulation in which participant controllers were exposed to one of two air traffic control situations within the operational envelope and their responses to a surface safety alert were measured. The results of this HITL simulation indicated the right tail-end of controller response behavior (defined as the time it takes a controller to detect the alert, make a decision about how to resolve the situation, and issue corrective instruction to a pilot) as around 13 seconds, which is almost twice as long as was measured in previous studies capturing controller response behavior in typical air traffic control situations.

Tower Controllers' Response Behavior to Runway Safety Alerts

As decision support tools to improve runway safety are introduced into Airport Traffic Control Towers, it is critical to understand their impact on controller performance. A Tower Human-In-The-Loop (HITL) simulation was conducted to evaluate the behavior of Local controllers in the presence of tower-based runway safety alerts. The results of this study suggest controllers like to gather as much information about the situation, within a reasonable amount of time (usually less than 5 seconds), before they initiate the action of issuing instructions to the aircraft involved in the emergency.