10
Conclusions and Recommendations

AUTOMATION ISSUES AND EMERGING TECHNOLOGIES

Levels of Automation

The levels of automation of any system vary along three dimensions: (1) information acquisition and integration (information automation), (2) decision and action selection, and (3) action implementation. The level of information automation is determined by the presence or absence of computer functions enabling filtering, information distribution, information transformations, confidence expressions, integration checks, and flexible information offerings based on the requests of users. Systems that possess all of these features have high levels of information automation, those possessing some have intermediate levels, and those possessing none have low levels. Automation of decision and action selection refers to the extent to which the controller's decision and action choices are constrained. Systems that have no or few constraints have low levels of automation on this dimension. Those that impose many constraints on operator selection of decision and action choices have high levels of automation of decision and action selection. Action automation refers to the actual implementation of an action choice and has only two levels: manual or automated.

RECOMMENDATION 1: The panel recommends that automation efforts focus on reliable, high level automation applications for information acquisition, integration, and presentation and for aiding controller decision making in order to support all system functions. Especially important in the



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 243
The Future of Air Traffic Control: Human Operators and Automation 10 Conclusions and Recommendations AUTOMATION ISSUES AND EMERGING TECHNOLOGIES Levels of Automation The levels of automation of any system vary along three dimensions: (1) information acquisition and integration (information automation), (2) decision and action selection, and (3) action implementation. The level of information automation is determined by the presence or absence of computer functions enabling filtering, information distribution, information transformations, confidence expressions, integration checks, and flexible information offerings based on the requests of users. Systems that possess all of these features have high levels of information automation, those possessing some have intermediate levels, and those possessing none have low levels. Automation of decision and action selection refers to the extent to which the controller's decision and action choices are constrained. Systems that have no or few constraints have low levels of automation on this dimension. Those that impose many constraints on operator selection of decision and action choices have high levels of automation of decision and action selection. Action automation refers to the actual implementation of an action choice and has only two levels: manual or automated. RECOMMENDATION 1: The panel recommends that automation efforts focus on reliable, high level automation applications for information acquisition, integration, and presentation and for aiding controller decision making in order to support all system functions. Especially important in the

OCR for page 243
The Future of Air Traffic Control: Human Operators and Automation near future is the development of decision aids for conflict resolution and maintaining separation. These aids should be directed primarily toward ensuring proper spacing between aircraft in preparation for the final stages of approach to landing and toward en route flight path efficiency improvement. RECOMMENDATION 2: The panel recommends implementation of high levels of automation of decision and action selection for system tasks involving relatively little uncertainty and risk. However, for system tasks associated with greater uncertainty and risk, automation of decision and action selection should not proceed beyond the level of suggesting a preferred decision/action alternative. Any consideration for automation above this level must be designed to prevent: loss of vigilance, loss of situation awareness, degradation of operational skills, and degradation of teamwork and communication. Such designs should also ensure the ability to overcome or counteract complacency, recover from failure, and provide a means of conflict resolution if loss of separation occurs. RECOMMENDATION 3: The panel recommends that the choice of manual (operator initiated) or automatic action implementation be guided by the level of automation of decision and action selection. Manual (or vocal) implementation is advised at the higher levels of automation of decision and action selection, at which automation narrows the decision action alternatives to a few, and more particularly at the level of automation of decision and action selection at which a single preferred decision/action is suggested. This manual (vocal) implementation will encourage the operator to review the contents of the recommended decision. RECOMMENDATION 4: The panel recommends that the availability of computer technology not be a reason for automation in and of itself. Clear requirements for functionality that can be achieved only by computer technology should drive design choices. RECOMMENDATION 5: The panel recommends that the choice of what functions to automate be guided by recognizing human strengths and the need to compensate for human vulnerabilities. Adaptable Automation Adaptable automation can benefit system performance by providing for the regulation of operator workload, reduction of complacency, and maintenance of manual skills.

OCR for page 243
The Future of Air Traffic Control: Human Operators and Automation RECOMMENDATION: If high degrees of automation of decision and action selection are to be introduced, adaptable automation should be considered so as to allow users to tailor the degree of automation to current needs and workload. Recovery Automation may increase capacity, but it will also increase traffic density and may increase airspace complexity by inducing changes in traffic flow from standard air routes. We conclude that continued use of automation of most functions eventually risks degradation of manual skills of operators who perform those functions. As a result, operators are likely to react more slowly to emergencies if they require use of those manual skills. And to the extent that the system requires manual recovery, those skills need to be preserved through recurrent training. To the extent that alternative skills will be required in emergencies, these new skills should be practiced and trained. Although automation may be highly reliable, either the automation or a resource on which it depends can be expected to fail or degrade at some time. Failure or degraded performance may come from software bugs, poor design, or aging hardware. Recovery response time will be greatly modulated by individual differences, redundant characteristics of the team environment, the complexity of the airspace (number of response options), and the familiarity versus the novelty of procedures necessary to cope with a degraded system. All important safety consequences of these failures are related to the margin by which available time exceeds the recovery response time. RECOMMENDATION 1: The panel recommends investing sufficient resources in studies of human response to low-probability emergencies. First, studies should be designed (1) to measure human response time (and accuracy) to improbable events, (2) to determine how to extrapolate to operational situations, using data on response times observed in experimental simulations in which it is known that a low-probability event could happen, and (3) to determine how response times are modulated by skill level. These studies should be conducted in the context of a specific system architecture and specified procedures for emergency recovery and system restoration. Second, failure modes/fault tree analyses should be actively pursued, particularly to identify situations in which two or more coordinating agents receive information inputs that are incongruous or contradictory. These analyses should be conducted on specific designs as part of the validation and verification process. Finally, human factors specialists should be involved in the development and testing of system recovery procedures. A research simulation facility should be available to these specialists for the study of human response to rare, unanticipated system events in the national airspace system.

OCR for page 243
The Future of Air Traffic Control: Human Operators and Automation RECOMMENDATION 2: The panel recommends the development of models, for given designs and procedures, to examine the implications of recovery in a high-density, unstructured airspace created by increased capabilities of ground-based automation or free flight. RECOMMENDATION 3: The panel recommends the development of airspace safety models that can predict the likelihood of midair collisions, as a function of the frequency and parameters of near-midair collisions and losses of separation, for varying standards of traffic separation. This will enable better prediction of the safety implications of capacity-increasing automation tools. RECOMMENDATION 4: In order to support airspace safety models, models should be developed that are sensitive to loss of situation awareness and possible degradation of skills that may result from moving operators to progressively higher levels of automation of decision and action selection. These models should be elaborated to incorporate compensatory gains that can be achieved by appropriate workload reductions and better integrated information. RECOMMENDATION 5: The panel recommends that air traffic control subject-matter experts collaborate with specialists in the behavioral sciences to model individual and team response to emergency situations and to populate the models with data to be collected in studies of human response time to low-probability emergencies. Policy makers should be made aware that choosing median response times to model these situations can have very different implications from those based on worst-case (longest) response times; these kinds of modeling choices must be carefully made and justified. RECOMMENDATION 6: The panel recommends that system functionality should be designed so that failure recovery will not depend on skills that are likely to degrade. Locus of Authority Future airspace projections dictate a need for increases in capacity without sacrificing safety. Two alternative vehicles for accomplishing these goals have been proposed: a free flight scenario and a scenario involving ground-based authority; both presume automation. Several different design concepts for free flight have been proposed, which vary in the degree of authority over control of the flight path allocated to airborne and ground systems. Those versions of free flight that assume high levels of airborne authority have the predicted capability for greatly increasing airspace flexibility and hence potentially increasing capacity

OCR for page 243
The Future of Air Traffic Control: Human Operators and Automation as well. However, a large number of uncertainties are associated with safety. These include uncertainties as to how pilot-to-pilot negotiations will be resolved in worst-case scenarios; problems relating to controllers' maintaining awareness of the tactical situation in an airspace made more complex and dense by the implementation of free flight; the workload impact of both increasing decision load in the cockpit and increasing monitoring load on the ground; and issues regarding possible confusion in the residence of authority among air traffic controllers, pilots, and airline operations center personnel. A ground-based alternative can incorporate the best features of projected air traffic control automation functionality related to interactive planning tools, conflict probes, and decision aids, all deployed with sensitivity to human-centered automation. It assumes greatly improved surveillance and communications bandwidth and accuracy. It also assumes greater availability of direct (user-preferred) routes than are currently available as a means to improve efficiency but maintains some consistency among those routes in order to limit airspace complexity and thereby support the controller's mental model of the airspace and facilitate failure recovery. Automated tools will enable negotiation of route changes as necessary. RECOMMENDATION 1: A ground-based scenario consistent with formulated plans of the Federal Aviation Administration can increase efficiency without radical changes in authority structure from the current system (e.g., the expanded national route program). The panel therefore recommends the development and fielding of current and proposed automation tools for ground-based air traffic control following the guidelines specified in this report regarding the selection of levels of automation. We also recommend the vigorous pursuit of projections of how various tools will operate in concert. RECOMMENDATION 2: Because free flight design concepts that assume a high level of airborne authority over control of aircraft flight paths have more uncertainties than design options involving ground-based authority with increased automation, the panel recommends extreme caution before existing levels of free flight are further expanded to greater levels of pilot authority for separation. Furthermore, we recommend the conduct of extensive human-in-the-loop simulation studies and validation of human performance models before decisions are made regarding the further implementation of free flight; this is needed to obtain reliable prediction of the safety implication of worst-case scenarios. We also recommend heavy reliance on scenario walk-throughs and focus group sessions with controllers, pilots, traffic managers, and airline dispatchers.

OCR for page 243
The Future of Air Traffic Control: Human Operators and Automation Teamwork Interpersonal communications and decision making between controllers and aircraft to resolve potential conflicts will continue to be an important component of air traffic control; it may assume greater importance under conditions of greater freedom in flight path choice. RECOMMENDATION 1: The panel recommends the continuation of formal training for controllers in teamwork, communications, distributed decision making, conflict resolutions, and coordinated response to unexpected events as a central aspect of controller training. Additional training for supervisors in interpersonal work skills should be a part of training and qualification. Automation of information representation and distribution has the capability to greatly facilitate teamwork between remote operators, by supporting shared situation awareness. RECOMMENDATION 2: The panel recommends the active pursuit of efforts to share dynamic information graphically among the various affected participants in the national airspace. Cross-Cultural Issues Research has demonstrated that there are large national differences in attitudes about and reliance on automation. Such differences may influence interactions between air traffic control and pilots of foreign air carriers flying in U.S. airspace. RECOMMENDATION: The panel recommends continuing to examine differences among nations in automation use in accordance with the recommendations of the Federal Aviation Administration's human factors team report on flight deck automation. Pilots from other nations that operate in U.S. airspace should be included in user tests of air traffic control automation. Emerging Technological Resources Visualization and Remote Control Computer graphic displays help visualization by combining variables into a single integrated display. The digital representation of altitude on the radar display has remained a feature of the air traffic control workstation that is less than optimal. Although controllers can adequately handle digital flight-level

OCR for page 243
The Future of Air Traffic Control: Human Operators and Automation data, it is difficult to visualize vertical trends from such a representation. One way of representing the vertical dimension in an analog format is through a vertical profile display; the other is through a perspective display. To date, the ambiguity associated with perspective displays remains a limitation for real-time air traffic control but offers promise for training. Intelligent Decision Aids Development of most decision aids requires a time-consuming and labor-intensive knowledge acquisition phase. Systems that learn may eventually reduce this bottleneck. However, at the present time, learning systems are not far enough along for operational use. Intent inferencing systems appear promising, but evaluations are needed of situations with more airspace complexity. Computer-Supported Cooperative Work Computer-supported cooperative work uses groupware technology to facilitate coordination, communication, and collaboration in accordance with the users' organizational and social context. New questions are raised about how to take social context and social process into account effectively when designing systems. To date there has been little systematic effort to apply this technology to time-critical operations such as air traffic control; however, there are some promising areas in which this approach may be useful, including strategic activities of air traffic management and interactions among tower controllers, airport managers, gate managers, pilots, and airline dispatchers in the surface movement advisor system. CURRENT AND ENVISIONED AUTOMATED SYSTEMS Surveillance and Communication Surveillance and Information Acquisition In order to maintain reliable performance, the radar processing system includes redundant equipment and is backed up by paper flight progress strips. Ongoing efforts to modernize aging equipment are expected to lead to systems with greater reliability. An alternative technology, the global positioning system, offers a high degree of accuracy; however, questions of its reliability, availability, and accuracy still need to be addressed. Satellites can be damaged, causing a hole in the constellation; satellites are always moving, causing receivers to change satellite sets to keep the necessary number (four) in view to establish an accurate position

OCR for page 243
The Future of Air Traffic Control: Human Operators and Automation (this can result in sudden steps in system error); jamming and spoofing are major concerns. The current distribution system of weather information for air traffic control is fragmented and does not adequately tailor information for controllers, traffic management specialists, and pilots. The key challenges are to provide additional useful weather information, integrate information from multiple sensors, predict weather more effectively, and disseminate information more efficiently to controllers, traffic management specialists, and pilots. Communications: Automatic Dependent Surveillance-Broadcast Mode Effective air traffic management depends critically on the accurate and timely exchange of information between ground and air and, increasingly, between aircraft. Considerable advances are derived from a communication system that can broadcast digital data, in parallel, to a broad range of airborne and ground-based users. Because automatic dependent surveillance-broadcast mode (ADS-B) provides an increase in both frequency and amount of information, it supports two potential expansions of the national airspace. First, it can potentially serve air traffic control with precise position information, thereby eventually replacing the slower, less accurate, and more expensive secondary surveillance radar. Second, the higher update rate and accuracy that ADS-B provides may enable more complex flight path negotiations between aircraft than does the present TCAS system. ADS-B is a likely enabling technology to support free flight. Communication: Data Link The use of visual and manual channels in data link substantially alters the process of communications, compared with traditional voice channels. The introduction of data link has very profound potential implications for the overall structure of the national airspace system and for the relationship among pilots, controllers, and automation. At one extreme, it is possible to envision a scenario in which humans, both on the ground and in the air, are substantially removed from the control loop, while control is exercised between computers on the ground and in the air. Although planners do not currently intend such a scenario, the possibility nevertheless exists that levels of automatic control and gating messages could be implemented that approximate this kind of interaction. This scenario leads to the high probability of loss of pilot awareness of the message content. RECOMMENDATION 1: The panel recommends the following approaches to ensuring redundancy in data link transmissions:

OCR for page 243
The Future of Air Traffic Control: Human Operators and Automation Provide redundant means of transmitting information contained in data link messages along conventional voice (radiotelephone) channels. Data link messages should be used primarily for routine communications (e.g., standard clearances, airport terminal information services). Radiotelephone channels should be reserved for the more unusual instructions and requests and for high-priority messages in high-workload (e.g., terminal) areas. Employ redundant voice synthesis of up-link messages as a design option, operated in parallel with visual (text and graphics) display of the message. RECOMMENDATION 2: The panel recommends the following approaches to defining the roles of flight crews and controllers in data link communications: Carefully analyze the possible role shifts and workload redistribution between personnel on the flight deck and between controllers at the workstation caused by data link. Training or design features should be used to address these role shifts if they are found to occur. Up-linked messages that directly pertain to aircraft control should not be automatically uploaded into the flight management system. Loading must be accomplished by an active choice by the pilot. This recommendation is consistent with our general recommendation concerning the need for careful evaluation of applying automation to high level system tasks. Flight Information Flight Management System The flight management system gives the pilot sophisticated, highly reliable tools to manage flight path control and power plant control with great precision. But with these ingenious tools have come problems at the human-computer interface, resulting in some degree of mistrust, overtrust, or mode confusion on the part of the pilots and, in the extreme, some spectacular incidents and accidents. It is essential that the same mistakes not be made in the implementation of the next generation of air traffic management systems. One of the problems that must be confronted is the incompatibility between the new flight management system aircraft and the constraints of the current air traffic control system. The full potential of the flight management system cannot be exploited in today's air traffic control environment. RECOMMENDATION 1: The panel recommends that the development of automation of air traffic control account for the capabilities of the flight

OCR for page 243
The Future of Air Traffic Control: Human Operators and Automation management system and, to the extent that safety is not compromised, be harmonized with those capabilities. RECOMMENDATION 2: The panel recommends that the lessons learned from the flight management system regarding mode errors and mode confusions be carefully applied to the design of air traffic control automation. Flight Data The design and implementation of electronic flight strips has been seen as a major risk to user acceptance in system automation. However, the electronic flight strip is only one means of modernizing the processing and display of flight data. The issue that needs to be addressed in the research and development process is less one of perpetuating the current roles and functionality of paper strips than of how to achieve an effective electronic embodiment of flight data. An electronic format for computerized flight information will facilitate the distribution of flight data and contribute to the reduction of controller workload. RECOMMENDATION 1: To facilitate operational acceptability of electronic flight information to replace paper strips, the design requirements should: Compensate for the redundancies provided by paper flight strips; Recognize how the characteristics of the paper strips (and procedures associated with them) support the cognitive processes of the controller; Develop a rapid and simple means of data entry; and Fully integrate flight data in an electronic work environment. RECOMMENDATION 2: The panel recommends that research studies undertaken to validate concepts for the integration of electronic flight data represent enough aspects of operational environments to allow for generalization of the results across operational settings. Immediate Conflict Avoidance Traffic Alert and Collision Avoidance System Although the traffic alert and collision avoidance system (TCAS) was originally intended to be a purely air-based system, designed to be a final backup to breakdowns in ground-based control, it is evident that it has much more profound implications for air traffic control. These implications will grow, as the system is extended to recapture more elements of the cockpit display of traffic information,

OCR for page 243
The Future of Air Traffic Control: Human Operators and Automation in the implementation of low levels of free flight (e.g., the role of TCAS in approaches; the role of TCAS in oceanic in-trail climbs). It appears that considerable thought was given to human factors issues in the initial implementation and subsequent fielding of the system. However, more early attention could have been given to trying to discover the complex pilot-controller interactions that have emerged, and that have subsequently forced revision of procedures, policy, training, and software. It is likely that more extensive reliance on system models (with valid models of human components), as well as complex human-in-the-loop simulation, could have anticipated some of these problems. It is encouraging to see movement in this direction as other future air traffic control technologies are envisioned. RECOMMENDATION 1: The panel recommends more effective training for pilots in order to foster greater consistency in response to TCAS alerts. RECOMMENDATION 2: The panel recommends that communication within the system be comprehensive in the sense that the information held by the controller with respect to neighboring traffic is accessible in the cockpit, and data on trajectory change instructions initiated by TCAS resolution advisories are electronically available and can be displayed to the controller as needed. Converging Runway Display Aid The converging runway display aid (CRDA) is a useful subsystem for TRACON operations at terminals where arrivals are directed to either one of two converging runways during normal operations. RECOMMENDATION: The panel recommends that (a) the methodological experiences with the converging runway display aid, including site adaptation procedures, should be used to inform the introduction of other new, special-purpose subsystems in the evolving national airspace system and (b) the mode of slot assignment and separation maintenance used by the converging runway display aid should be considered as a possible benchmark in the design and refinement of alternative subsystems for terminal area operations. Precision Runway Monitor The precision runway monitor/final monitor aid system has generally good user acceptance. However, some highly experienced controllers have voiced reservations about the passive monitoring role of the system operator. Even with runway separation distances reduced to 3,000 feet between dual runways (as

OCR for page 243
The Future of Air Traffic Control: Human Operators and Automation approved in November 1995), the frequency of transmissions may be too low to allow the controller to sustain a reasonable level of alertness. RECOMMENDATION: The panel recommends that (a) studies be conducted to determine whether the problem of vigilance decrement can be avoided by the integration of the precision runway monitor/final monitor aid system with the approach control system and (b) the trade-offs between ground-based and cockpit-based systems for lateral separation be carefully considered. If redundant systems are implemented in the air and on the ground, all possibilities by which conflicting guidance from the two systems might be given should be analyzed (e.g., because of different sensors, different conflict prediction algorithms, different communication bandwidths). Avoiding Collisions on the Ground To address both safety and efficiency concerns, the Federal Aviation Administration is undertaking a set of activities that, taken together, are intended to provide controllers and pilots with automated warnings of potential and actual runway incursions and ground traffic conflicts, with automated means of communication and with the capability to maintain situation awareness in low-visibility conditions. These initiatives range from current implementation, through near-term enhancement, to long-term development programs. The combination of automated functions can potentially introduce effects that are not predicted from studies or tests of each automated function independently. Also at issue is the distinction between trust in the system and trust in its components. Individual components may vary in their trustworthiness (e.g., ASDE radar and GPS/ADS-B), and a thorough understanding of the capabilities of each component, as well as how the components work together (e.g., AMASS display and runway status lights), is required to permit pilots and controllers to develop an appropriate level of trust in the system. In addition, since these new systems are specifically intended as safety enhancements and may also be used to support increased usage of airport surface capacity, it is particularly important that controllers and pilots are able to respond effectively to failures (e.g., degradation of sensors or sensor integration software). RECOMMENDATION: Because a variety of ground-based and aircraft-based sensors and information processors are envisioned, the panel recommends that a careful analysis of failure modes and effects for the total system be undertaken to ensure that conflicting information is never provided to pilots and controllers regarding the status and safety of runway and taxiway paths. Controllers and pilots should receive specific training that allows their understanding of system functions and limitations.

OCR for page 243
The Future of Air Traffic Control: Human Operators and Automation Strategic Long-Range Planning Center TRACON Automation System The organizational implications of the center TRACON automation system remain uncertain. A strength of the system is that it is designed to be advisory only. Therefore, by not directly affecting required procedures, the potentially negative impact on organizational functioning should be minimized. It derives useful advice for the controller that without the system would be cognitively difficult to derive. It also facilitates sharing of information between controllers. There is a possibility that extensive reliance on CTAS could create an airspace that is denser and more complex, creating higher levels of controller perceptual workload. RECOMMENDATION: The panel recommends the active role of human factors resources at all stages of development of the center TRACON automation system, including ongoing field tests. In addition, the informed input of users should be secured in defining and refining the functionality as well as the interface. This process should be repeated in the fielding of other systems. Adequate and extensive training should continue to be given to users regarding the assumptions underlying the system's advisories. Conflict Probe and Interactive Planning Although experienced controllers have developed considerable cognitive skill in predicting aircraft trajectories, additional tools for controllers may facilitate this skill. Conflict probe and interactive planning tools are designed to support more long-range strategic planning and to address human visualizations in this regard. They represent appropriate higher levels of automation of information gathering and low levels of automation of response. Both tools have the potential, depending on specific design characteristics and associated procedures, to permit reallocation of control tasks between the R-side and the D-side controllers. For failure recovery the primary issues are twofold. It is possible to envision a scenario based on one set of design characteristics in which an interactive planning tool—the user request evaluation tool or the conflict probe system—first has enabled more complex (and possibly more densely packed) traffic flow, enabling user-preferred trajectories and, second, has left the R-side or tactical controllers with reduced situation awareness of the current airspace (because changed trajectories were not imposed by their decisions). A sudden failure within the system could leave the tactical or R-side controller more vulnerable in issuing the rapid tactical commands necessary to avoid conflict situations.

OCR for page 243
The Future of Air Traffic Control: Human Operators and Automation RECOMMENDATION: Simulations should be conducted to examine traffic complexity levels that are generated by the use of interactive tools to grant user-preferred trajectories. The system should be introduced in such a way that all controllers on a team can maintain full situation awareness of routing changes supported by the tool. Four-Dimensional Contracts Four-dimensional contracts will change aspects of the controller's job and are likely to create a less well-structured, more densely packed airspace, but they will not fundamentally alter responsibility for separation. RECOMMENDATION: The panel recommends closely following European demonstration projects concerning the usability of four-dimensional trajectory planning and flight path negotiation tolls for lessons learned and potential application in the United States. Surface Movement Advisor The surface movement advisor (SMA) is intended to provide controllers, pilots, airfield managers, ramp operators, and airline operations personnel with automated support of surface traffic planning. Airport area automation holds the potential for changing the roles of controllers vis-à-vis pilots and airport and airline personnel. Realignment may include new responsibilities, new authority structures, new communication and cooperative work links, and new measures of effectiveness (e.g., increasing emphasis on efficiency). Since a prerequisite for its design is data distribution through computerized networks to cooperating team members, one promising avenue that can contribute to the design of an effective surface movement advisor is the combination of analyses and tools that pertain to emerging computer-supported cooperative work technology. RECOMMENDATION: The panel recommends that (a) the impact of the surface movement advisor on individual roles and on teamwork be carefully analyzed during analysis, design, and test activities and (b) computer-supported cooperative work analyses of the surface movement advisor be performed. Related tools should be applied when analyses deem the applications appropriate. Support Functions The planned centralization of maintenance activities and the projected trend toward automating more complex cognitive functions now performed by maintainers represent a fundamental shift for maintenance design, operations, and

OCR for page 243
The Future of Air Traffic Control: Human Operators and Automation organization. The skills of the maintainers are a significant factor in system integrity. The recently developed GS-2101 classification, increasingly applied to airway facilities specialists, outlines requirements for systems engineering skills (as opposed to component or subsystem maintenance skills) needed to support this shift. These changes are proceeding amidst a paucity of knowledge regarding: (1) maintainer task performance and error while interacting with highly automated systems, (2) the mental models of the system that guide maintainer decisions and actions, and (3) the variables of teamwork—both among maintainers and between maintainers and air traffic controllers—involved in system monitoring, control, and maintenance. In addition, the GS-2101 classification has proceeded without supporting development of validated selection, training, and performance standards for the anticipated systems engineering task requirements. RECOMMENDATION 1: The panel recommends that representations of maintainers' mental models be developed to complement cognitive task analyses for maintainers. These models and analyses, as well as human factors principles, should be used to develop a reasoned approach to automation and to the design of new maintainer's workstations, especially for centralized operations control centers and the national maintenance control center. RECOMMENDATION 2: Selection, training, and performance standards should be developed and validated appropriate to the knowledge, skills, and abilities required to maintain highly automated systems. RECOMMENDATION 3: The maintenance teamwork and the coordination between maintainers and air traffic controllers should be examined in the context of new centralized operations control centers. The increased role that controllers may assume in maintenance tasks, given digital technology of automated systems, should be considered. RECOMMENDATION 4: The reliability of maintainers in automation-supported maintenance tasks should be studied, and an error-tolerant design should be applied to maintenance equipment. INTEGRATION The Future National Airspace System Authority is a critical concept in the evolution of the national airspace system, whether this evolution is toward a concept of free flight or ground-based automation.

OCR for page 243
The Future of Air Traffic Control: Human Operators and Automation RECOMMENDATION: The panel recommends that the momentary residence of authority with controller, an automated agent, or the pilot is unambiguously announced and displayed to all relevant agents, especially in systems in which this authority may shift dynamically. It is important to realize that the concept of free flight remains somewhat ill defined. Different players have very different notions of what it should be, how free it should be, and over what domains of the airspace it should apply (e.g., en route versus TRACON, high altitude versus all altitudes, continental versus oceanic). One of the contentious issues to be addressed regarding free flight concerns the appropriate level of authority that should be maintained by air traffic controllers in a free flight regime. An airspace that functions under free flight rules will lose the structured order that enables the controller to easily grasp the big picture. It is quite possible that an airspace under free flight will yield unpredictable shifts in traffic density, and this in turn may require some degree of ''dynamic resectorization." Finally, free flight separation algorithms, like those of TCAS, are likely to be time-based rather than space-based. Space can be easily visualized by the controller, but time less so. It is unclear the extent to which this shift may also inhibit controller situation awareness. One of the greatest challenges is to try to predict the implications of changes in free flight procedures to overall system safety. RECOMMENDATION 1: The panel recommends three parallel research approaches to estimate safety implications: Continue to refine simulation and modeling with an emphasis on modeling safety parameters. Collect sufficient amounts of human-in-the-loop data to populate simulation models that can be used to identify, understand, and compensate for infrequent (but catastrophic) consequences. The results should provide the basis for understanding responses to such events under conditions that approximate realistic occurrences. Rely heavily on scenario walk-throughs and focus group sessions among controllers and pilots who have been provided with clear descriptions of future assumed capabilities; these can reveal potential bottleneck areas. RECOMMENDATION 2: The panel urges exploration of design changes that offer the possibility of greatly increasing the safety margin between aircraft, even as the procedures are altered to allow more regular flow (i.e., improve efficiency). Examples of design changes include satellite navigation, ADS-B communications, and automated tools for medium- and long-range conflict probes.

OCR for page 243
The Future of Air Traffic Control: Human Operators and Automation Introducing Automation Development and Installation of Advanced Systems The introduction of automation, whether incremental or comprehensive, involves some interference with an ongoing process that cannot be disrupted. Consequently, careful planning is required so that the transition can be made with minimum interruption. Despite the FAA's management efforts to foster greater human factors involvement in the development and implementation of advanced air traffic control systems, the agency's success record has been mixed at best. However, a recently completed, FAA-commissioned, independent study (by the Human Factors Subcommittee of the FAA's Research, Engineering, and Development Advisory Council) examined the current FAA organizational structure, staffing, and operating practices as they relate to human factors support activities, and made recommendations for improving the effectiveness of this function. These recommendations appear to be well founded and offer the potential for better integration of human factors concerns in the development of advanced automation technologies. RECOMMENDATION 1: The panel recommends that senior Federal Aviation Administration management should reexamine the results of the study by the Human Factors Subcommittee of the FAA's Research, Engineering, and Development Advisory Council, with a view toward implementing those recommendations that appear most likely to achieve more active, continued, and effective involvement of both users and trained human factors practitioners. All aspects of human-centered automation should be considered in fielding new automated systems. RECOMMENDATION 2: The Federal Aviation Administration should continue to support integrated product teams with well-trained human factors specialists assigned to the teams. Both users and human factors specialists should be involved at the early stages to help define the functionality of the proposed automation system. These specialists should be responsible to report to human factors management within the Federal Aviation Administration as well as to project managers. RECOMMENDATION 3: The Federal Aviation Administration should continue to work toward an infrastructure in which some human factors training is provided to personnel and program managers at all levels of the organization (and contract teams).

OCR for page 243
The Future of Air Traffic Control: Human Operators and Automation RECOMMENDATION 4: The Federal Aviation Administration should ensure that adequate funding for human factors work is provided at all stages of system development and field evaluations both before and after systems acquisition. RECOMMENDATION 5: Starting with early conceptual development and continuing through installation, a system-specific analysis should be undertaken of the interactions between system attributes and operators' capabilities. Implicit interdependencies among controllers and between them and other human operators or automated agents should be taken into account. RECOMMENDATION 6: Contextually valid controller-in-the-loop experiments and simulations should be conducted to validate, test, and refine system design. Human factors professionals should advise in the conduct of these experiments, with attention to good experimental design and adequate sample size. Organizations such as the Federal Aviation Administration Technical Center, the Civil Aeromedical Institute, and NASA should remain heavily involved in these developmental efforts. RECOMMENDATION 7: The panel recommends proceeding gradually with the introduction of automated tools into the workplace, giving adequate attention to user training, to facility differences, and to user requirements, and carefully monitoring the operational experience from initial introduction, putting mechanisms in place to respond rapidly to both positive and negative lessons learned from those experiences. RECOMMENDATION 8: The panel recommends that operators chosen to work with new systems or subsystems should be given an understanding of the principles of system operation, including the logic and algorithms underlying the system as well as the practice of system operations. Training should progress quickly to the level of real-time exercises in the setting of interactive simulations. Embedded training should be considered as a useful approach to help controllers maintain skills. Valid and reliable performance measures should be developed and proficiency should be defined with regard to the specific measures. Long-Range Planning The pace of events is such that some advanced subsystems, such as the converging runway display aid, are already installed at selected locations, and others, such as the center TRACON automation system, are on the verge of operational installation. Meanwhile, system-specific studies are under way on other systems such as the global positioning system and data link. Now is the

OCR for page 243
The Future of Air Traffic Control: Human Operators and Automation time for the design and implementation of studies that deal with some of the generic problems of air traffic control and advanced technology, such as the effects on system performance of passive monitoring by controllers using the precision runway monitor. Only by building the knowledge base now will the FAA be able to make sound decisions about the future cycles of automation and to help eliminate surprises from each successive wave. RECOMMENDATION 1: During development of each automation function, system developers should consider possible interactions with other automation functions (under development or already existing), tools, and task requirements that form (or will form) the operational context into which the specific automation feature will be introduced. RECOMMENDATION 2: Various research methods should be integrated: models, high- and medium-fidelity simulations, and more controlled laboratory experiments at all levels of system development. Laboratory experiments that can address many useful questions of interface design must consider contextual relevance. Results of these experiments should be used to inform more realistic simulations about what variables should be investigated. The results from these experiments should be used to help estimate and validate human (pilot and controller) performance parameters for computational models.