Philip J. Smith
Charles Billings
Amy Spencer

Cognitive Systems Engineering Laboratory

Institute for Ergonomics

The Ohio State University

September 30, 2001

This report was produced under a subcontract from Raytheon to complete work for NASA’s Advanced Air Transportation Technologies program. It was made possible through cooperation with the Airline Dispatchers Federation.

This report summarizes findings based on a review of the literature on human factors issues in the functioning of Airline Operations Centers (AOCs) and a series of interviews with senior dispatchers from 5 airlines. Nine general hypotheses were generated from this data collection effort:

• NASA’s tool FACET offers significant promise both as a simulation environment to study traffic flow management strategies of interest to AOCs, and as a tool to be linked with airline scheduling and flight planning systems in order to provide forecasts of potential traffic bottlenecks both for pre-flight planning, and for reevaluation of these plans while flights are enroute. To more effectively support these applications, FACET needs to incorporate stochastic models of departure (off) times, thus making it possible to complete sensitivity analyses by reasoning about uncertainty.

Whether the supporting tool is based on some future version of FACET or some other approach, the goal should be to provide dispatchers with information about predicted traffic constraints, and then allowing them to take the initiative to plan around those traffic constraints rather than having departure delays or assigned reroutes mandated by the Air Traffic Service Provider (ATSP). (The same paradigm was recommended for dealing with predicted weather constraints, allowing dispatchers to take the initiative to plan around weather rather than having the ATSP mandated departure delays or weather avoidance route.)

• NASA’s tool Direct-To provides an example of the potential value of tools to help the airlines reevaluate the flight plans for specific aircraft while they are enroute, reassessing flight plans in the context of changing traffic congestion and weather constraints as well as in terms of safety and airline business concerns. To ensure safety when using such tools, however, as well as to ensure consideration of airline business concerns, such tools and procedures need to be designed to incorporate dispatcher judgments as part of the decision making process. With this requirement in mind, it was suggested that, with appropriate refinements, Direct-To and other tools like it represent a potentially valuable approach for tactically adjusting flight plans to airlines to achieve their business objectives.
• The dispatchers interviewed indicated that, while it is important to develop new procedures and tools that provide the airlines and other system users with increased flexibility to meet their business concerns, in general this should not be done at the expense of significantly reducing system capacity when compared to alternative approaches to providing increased flexibility. Furthermore, regardless of the approach taken, any system enhancements should be achieved in such a way as to increase safety.
• The current National Airspace System (NAS) is a diverse environment where it is unlikely that specific solutions can be universally applied to improve performance in a cost-effective manner. It is therefore important to consider hybrid solutions that make use of a repertoire of alternative procedures and tools to match specific airport and airspace constraints. One useful distinction is to categorize the NAS in terms of areas with congested airspace (such as the Northeastern United States) vs. areas with relatively uncongested airspace.

Regarding the strategic application of unconstrained "free flight" in regions with congested airspace while enroute (allowing preflight filings and "significant"flight plan amendments while aircraft are enroute without any structuring of the overall traffic flows), the dispatchers interviewed judged that this is likely to result in reduced system capacity when compared with more structured traffic flows.

• In addition to providing expert judgments regarding approaches to strategic and tactical control in the enroute environment, the dispatchers discussed issues concerning Controlled Time of Arrival (CTA) as a parameter for control by the ATSP. There was general support for this concept, although some dispatchers viewed it as an interim step that should become less necessary if enroute, terminal area and surface capacities are appropriately increased in the future.

• As an alternative or complement to other strategies relying on preflight planning of routes and preflight assignment of CTAs, some dispatchers suggested exploring a more dynamic system that would reevaluate plans at wheels up or at top-of-climb for each flight.
• To increase safety and system capacity, and to provide the airlines with more flexibility to meet their business concerns:
• The roles of the dispatcher and other AOC staff will need to evolve to include new responsibilities for monitoring and responding to the impacts of traffic, weather and flight-specific situations that develop while flights are enroute;
• New support technologies and procedures will need to provide different system level perspectives, supporting congestion management and schedule management. Technologies and procedures focusing on the performances of individual flights need to be integrated into solutions based on these broader system level perspectives in order to achieve the desired benefits.
• In order to evaluate new approaches to strategic and tactical control in the enroute environment (such as strategic and tactical "free flight"), a number of critical scenarios were identified. Some of these focus on the impact of the control paradigm on system capacity. Others deal with safety issues. Still others are concerned with opportunities to better achieve airline priorities.
• A number of the conclusions above suggest the need for AOC involvement in simulations of future tool and system designs. The participating dispatchers suggested that it is would be more cost effective to develop a distributed simulation capability where NASA simulations could interact with AOC capabilities at participating airlines. This recommendation was based on the fact that simplistic approaches to simulating AOC functions (such as a workstation that focuses only on flight plans based on time and fuel considerations) are unrealistic, and are not likely to adequately represent AOC functions within such scenarios.

The purpose of this report is to provide preliminary results from research regarding Airline Operations Control (AOC) perspectives (Beatty, 1995; Lacher and Klein, 1993) on issues of interest to NASA. This report summarizes findings from two sources:

• The human factors literature on dispatch operations regarding different strategies for congestion management, flight planning and schedule management by the airlines (Braune, et al., 1996; Kerns, Smith, et al., 1999).
• 21 hours of structured interviews with 13 dispatchers. Six topics are covered in this discussion. These deal with AOC perspectives on:
• FACET, a simulation tool under development by NASA.
• Direct-to, a route analysis tool currently being tested by NASA.
• Approaches for incorporating dispatch functions into NASA human-in-the-loop simulation capabilities.
• Approaches to strategic control in the enroute environment.
• Approaches to tactical control in the enroute environment.
• Different classes of scenarios for evaluating alternative system or subsystem designs from an AOC perspective.

Within this discussion, a number of human factors considerations underly the assessments and recommendations (Krozel and Mogford, 2000) . These issues deal with:

• Helping individuals deal with the cognitive complexity and uncertainty associated with planning in the NAS through strategies for distributing responsibility and work, and through the introduction of decision support tools to help with reasoning and computation (Billings, 1997; Corker, et al., 2000; Orasanu and Salas, 1993; Rasmussen, et al., 1991; Robertson, et al, 1990; Sheridan, 1987; Sheridan, 1992; Smith, Billings, et al., 2000a, 2000b, 2000c; Smith, McCoy, Denning, et al., 1997; Smith, McCoy and Orasanu, 2001; Smith, McCoy, Orasanu, et al., 1997).
• Ensuring that, within the distributed work environment, each individual has an accurate and appropriate mental model of the situation as it applies to his or her responsibilities (Endsley, 1997; Hilburn, et al., 1998; Orasanu, 1991).
• Recognizing that current limitations are not due solely to complexity issues, but that there mental workload constraints that also limit system performance (Avans, et al., 1997; Hilburn, Bakker, et al., 1997; Hilburn, Pekka, et al., 1997).
• Identifying situations where contingency planning is needed.
• Understanding that "to err is human" and that there is a need for safety nets (effectively designed redundancies) to make sure that an error made by one person (by an operator within the system or by the designer of a tool used by an operator), will be caught and dealt with before an adverse outcome occurs (Smith, McCoy, et al., 1997; Weitzman, 1997).
• Ensuring that the allocation of control matches the distribution of data, knowledge, processing capacity and motivation (Beatty and Smith, 2000; Beatty, Smith and Billings, 2001; Smith, Billings, et al., 1997; Smith, Billings et al., 1998; Smith, McCoy and Billings, 2000).

For clarity, the stages of a flight have been divided into 4 segments in the following discussions (Hopkin, 1995):

• Pre-flight (prior to pushback from the gate).
• Departure (from pushback to top-of-climb).
• Enroute.
• Arrival (from top-of descent to arrival at the gate).

In addition, a distinction is made between tactical changes while enroute that return to the originally planned route (small deviations from the flight plan in order to avoid some localized constraint such as a storm cell or traffic that do not "significantly" impact traffic flows or airline schedules), and strategic changes while enroute ("significant" deviations that have a substantial impact on the time, fuel and/or airspace traversed by a flight, or that have a "significant" impact on traffic flows or on how that flight contributes to the airline’s schedule). Note that, as defined here, this distinction between tactical and strategic changes is not a simple one.

The complexity of this distinction is provided by the following example produced using POET from April, 2000. This example illustrates how, in some parts of the NAS, physically small deviations to avoid a storm can have a significant impact on traffic flows and airline schedules, thus making these deviations of strategic importance.

In this example, early in the morning on 4/6/2000 PHL west bound departures were shut off over PTW, and OOD was not an option due to the ZDC DSR transition procedures. J60, J64 and J80 were closed per ATC. There was weather over PSB and deviations by flights as a result of this weather. USAirways experienced a number of lengthy delays to and from PHL during this event.

The first figure below shows the weather in this timeframe. The second shows flights deviating to the north and south of the location of a storm cell, and flights put into no-notice high altitude holding due to the resultant traffic bottlenecks. (20% of the 40 flight instances that were scheduled to arrive into PHL in the 1200 Z hour experienced high altitude circular holding.) In the second figure, flown routes are shown in black and filed routes are shown in green.

As commented upon by a traffic manager from ZNY:

"When this happens because the pilots indicate they can’t fly through the weather, you don’t have much room to maneuver. With the size of the sectors in New York Center, once the aircraft start deviating around the weather it shuts down other routes."

The hypotheses developed as part of this study are based on a review of the human factors literature on AOC functions in the NAS, and on a series of structured interviews with senior dispatchers. The reference list at the end of this report contains citations for the pertinent literature.

In terms of the structured interviews, 6 hour and a half long telephone interviews were conducted. These interviews were conducted with 2-3 dispatchers at a time on a teleconference.

A total of 13 dispatchers participated in these telephone interviews. On average, these individuals had 16.2 years of work experience as dispatchers. This set of dispatchers included individuals with prior work experience as air traffic controllers, as air traffic control coordinators for airlines, and as chief dispatchers for airlines. It also included individuals who have been members of RTCA, FAA and NASA advisory committees and steering committees for the design of the future NAS, and who have been officers of the Airline Dispatchers Federation.

5 of the 13 dispatchers who participated in the telephone interviews then participated in a 12 hour focus group over a two day period. The goal of this focus group was to elicit additional details regarding AOC perspectives, and to provide input regarding the design of specific tools currently under development by NASA.

FACET (Future ATM Concepts Evaluation Tool) is a simulation that has been developed by NASA Ames Research Center to study questions dealing with airspace congestion. This tool can access a database with historical or current ETMS data and identify congested sectors. It can also be used to compare the impacts of alternative traffic management strategies (such as different strategies for applying miles-in-trail).

The five senior dispatchers who participated in the12 hour focus group were unanimously positive about the potential use of FACET for both simulation studies and as a tool for real-time use by AOCs, as illustrated by the following quote.

"If you want to expand your platform of test cases, let me know, because I would frankly love to see -- I would love to discover what this could do for me as I do my job, because it seems to have tremendous potential to help me do a better job, to help all of us do a better job as dispatchers."

"So FACET could come to me as a tactical air traffic coordinator for an airline, and tell me there's 10 flights going to leave Chicago in the next 30 minutes going over North Brook. Two of your flights are in those 10. One of those flights is going to take a 30-minute delay unless somebody off-loads to this other fix. Now I'm faced with a business, and safety decision that says do I flip the coin and risk the 30 minutes, or do I take seven minutes if I go this over way. I'll probably take the seven minutes sure delay. That kind of predictive information would be very useful."

"FACET could let me test a reroute, put that into the equation, and predict what would happen. If it could also help me identify a list of flights that are going to be impacted by some constraint, and suggest an optimized solution that affects the minimum number of flights, or affects a larger set of flights the minimum amount, that would help. It could auto-suggest not just one reroute, but several different reroute options to shortcut the dispatcher to a good solution. It would still be my place, though, to decide to move them."

"What would be useful to me would be, show me three or four routes that are valid, and that are available."

"Suppose I want to move flights down to the south, and I ask: "Could I just fly the route -- file the route here to begin with?" And the answer is: "Because you're not supposed to fly through Cincinnati's airspace." So, part of the issues in the long-term are identifying what are all the potential kinds of constraints, and how do you make it easy for a dispatcher or someone to say "These are the viable routes."

If you don’t know that, you're sitting on the ground. You can't go. Whereas if the dispatcher knew enough to develop a flight plan that avoids Indy Center, you could go right on by it."

"It would let us look at a miles-in-trails plan and ask: Do we need the miles-in-trail if I move those flights, and what is the delay I would get if I leave it the way I originally wanted it? The ground delay program model is the same exact model. We sit there when they issue a ground delay. We've got our EDCT's. We go in and model, and say: "If we cancel this one and this one, what happens?" It's click, click, like automated substitution, and we go: "Yeh, that's going to solve our problem."

Additional details are provided below.

There are a number of questions of interest from an AOC perspective that could be addressed using FACET as simulation environment. These include:

• Studying the impacts of different traffic flow management strategies, such as the use of Low Altitude Arrival and Departure Routes, the use of Playbook routes and the collaborative use of Coded Departure Routes.
• Evaluating the effectiveness of airline-initiated actions to deal with forecast traffic congestion or weather.

In order to address some of these questions, it would be necessary to develop new capabilities or datasets within FACET, such as the inclusion of a rule-base to simulate the flexible selection of alternative departure routes based on the rate of development of a weather pattern. Others would require the incorporation of a tool to allow AOCs to generate a database of proposed routes in response to a particular set of forecast weather or traffic constraints:

"I would definitely want to do a study research on how much the pre-emptive actions of the user community could solve the total pie of the air-traffic problem. Use the simulator, do human-in-the-loop testing with dispatchers on a variety of scenarios."

In addition to developing new capabilities to simulate scenarios of interest to AOCs, it is also important to identify appropriate measures of performance. Some of the measures of interest are already incorporated into FACET, such as sector loadings and the departure or airborne delays associated with particular flights. However, new metrics that look at impacts on different aggregations, such as the delays for flights filing a particular departure fix, also need to be identified and incorporated. Another example of a possible benefit to measure was:

At present, airline scheduling and flight planning tools function without access to any direct information or predictions about traffic bottlenecks. The modeling capabilities offered by FACET offer a potentially suitable approach for providing look-ahead during preflight planning and during enroute re-evaluations of flight plans.

The general strategy would be to link airline flight planning tools with FACET, allowing an AOC flight-planning and scheduling tool to submit proposed routings for a single flight or a collection of flights for evaluation against the current data in ETMS on filed routes. FACET could then provide feedback on predicted congestion along that route.

To accomplish this integration, there are a number of enhancements that need to be completed. These include development of:

• A communications protocol for interactions between FACET and airline flight planning tools.
• Stochastic models to represent the impacts of uncertainties about departure times in order to perform sensitivity analyses and complete more realistic evaluations of alternative routes using FACET.
• More sophisticated metrics of airspace saturation to accurately predict system performance.
• Visual representations that would allow dispatchers to view the predictions of FACET regarding airspace congestion as a function of the planned location of a flight over time.
• Automation to allow airline flight planning and scheduling tools to probe airspace congestion predictions from FACET in order to search for desirable routes and schedules.
• What-if capabilities that would allow AOC scheduling and flight planning tools to assess the aggregate impact of alternative flight plans for collections of flights.
• Re-evaluation algorithms in FACET that could determine the need to re-assess a flight plan when there is a significant change in either the performance of that particular flight (such as a delay in its departure time) or in the congestion of the airspace that it plans to traverse. (As an example, traffic managers have indicated that it is difficult to predict prior to departure whether a flight from Southern California to DFW that is filed north of White Sands will encounter saturated airspace near UKW because of uncertainties about its departure time and the departure times and routes of other flights filed into UKW.)

There was also one caution about too naively making such a linkage between airline flight planning and scheduling tools and FACET:

"On my part, I would be be careful about just thinking of plugging an airline flight planning system in for some of these folks, because the flight planning systems themselves have been de-optimized."

It is very important to note that underlying this suggested use of FACET is recommendation of a paradigm shift for how air traffic is managed. A tool like FACET is of interest to AOCs because they believe that a better way to manage traffic is to tell dispatchers where the constraints are and to let them flight plan around those constraints before having the ATSP step in and impose a solution (ground delays or reroutes). This proposed approach applies to both traffic and weather constraints:

"We need real-time information about predicted traffic loadings so we can plan around traffic."

"We’ve been asking for a common situation display for a long time. This may be the basis for it."

In order to use FACET as a simulation tool or as a real-time decision support tool in AOCs, work needs to be done to further test and enhance the accuracy and precision of its predictions. This requires the identification of an appropriate set of test scenarios and the development of a corresponding database representing the flights included in those scenarios.

Direct-To is a second tool under development and testing at NASA Ames. This tool is designed to re-evaluate the flight plan for an aircraft as it progresses along its route. Direct-To monitors for opportunities to avoid flying doglegs in a planned route if the airspace that allows a shorter direct route for that segment of the flight becomes available. Direct-To looks at traffic predictions along more direct routings and alerts the controller of opportunities to cut off a dogleg. The controller, with the concurrence of the flight crew, then decides whether such a recommendation is "desirable."

The dispatchers who participated in the focus group were supportive of tools like Direct-To that help in the re-evaluation of a flight plan while it is enroute. They strongly recommended, however, that such tools be designed to ensure that dispatch is involved in such decisions.

They recommended dispatch involvement at two points:

• Preflight (there are some cases where the dispatcher needs to indicate in the original filing that it is inappropriate for a flight to accept a more direct route along some flight segment. Ideally, this indication would provide a short explanation so that the flight crew understood the basis for this recommendation).
• While enroute.

All of the dispatchers agreed that dispatch needed to be involved in any decisions about direct routings that represent "significant" changes. This requirement is mandated by the FARs, and is of considerable importance in terms of both the safety and efficiency of a particular flight, as well as in terms of the impact of such a change on the airline’s overall schedule. There were some unresolved disagreements among the participating dispatchers as to whether dispatcher concurrence should be required for all such direct reroutings (even when they are very tactical in nature), or whether such concurrence should be required only when the impact represents a "significant" impact on the time, fuel and/or airspace traversed by a flight, or a "significant" impact on traffic flows or on how that flight contributes to the airline’s schedule. This issue (including the definition of "significant") needs further consideration.

Related to this is the issue of timing and workload. There will be cases where, to take advantage of a more direct route along some segment of a flight plan, the decision will need to be made quickly. This suggests the need for some type of look-ahead that would alert the dispatcher and flight crew well ahead of time regarding a potential opportunity, so that a considered decision can be made in a timely fashion.

One example emphasizing such concerns about dispatcher involvement is in cases where a more direct routing can result in a significant underburn of the fuel loaded for a flight, making it too heavy to safely land upon arrival at its destination. (This already sometimes happens in the current NAS when flight crews fail to consult Dispatch in situations where they agree to accept a more direct routing from a controller.) Other examples involve checking for compliance with MELs, and the need to consider the impact of a flight’s early arrival on sector and taxiway congestion and on gate assignments. (A good illustration of this is provided in the figure below, which shows a flight that was filed by the dispatcher over land, but that was amended to fly direct without dispatcher involvement. The direct route takes an aircraft that is not overwater-equipped over the Gulf of Mexico.)

More generally, the participating dispatchers indicated that tools like Direct-To that help re-evaluate a flight plan while an aircraft is enroute could have considerable value. They emphasized, however, that such decisions cannot be based on oversimplistic assessments of traffic congestion or weather that fail to incorporate the many factors considered by the dispatcher as part of the decision making process. This perspective is captured in the following statements by dispatchers.

"When he offers a more direct route, does the controller know if there is

"The dispatcher needs a way to indicate on a flight plan that a given flight should not accept direct. We could say something like "Captain, do not deviate from the filed route today if offered direct flight ATC. Unfortunately, we have no way of getting the pilot to read that. (Laughter.) We could put "Free beer at layover hotel" or something like that to tell if they're reading. But what we need is a way to tell the automation and the controller not to offer a direct in some cases."

This concern about having dispatchers in the loop is further emphasized by the following exchange.

NASA Researcher: "Over what?"

Dispatcher: "I can tell you this unequivocally, if that was a 727 that was full, and

Another example emphasizes the need for redundancy (keeping dispatch in the loop as well as the flight crew and ATC) as a safety net protecting against a slip or mistake made by any one person. In this case, an aircraft that was overwater equipped had to be replaced because of a mechanical. The original aircraft was overwater equipped and would have been filed over the Gulf of Mexico. Because the replacement aircraft was not overwater equipped, however, the dispatcher filed it on an overland route. After takeoff, the pilot accepted an amendment to fly direct from a controller that put them on an overwater route.

The dispatchers interviewed cautioned against simulating an oversimplistic view of the contributions of AOCs to the operation of the NAS. This caution focused on two areas.

The first area of emphasis was on the need to consider performance from a broader systems perspective in designing and conducting simulations. This means thinking in terms of the impact of new system designs on schedule management, congestion management and throughput rather than simply looking at changes in time and fuel consumptions for individual flights.

The second area of emphasis was on the need to evaluate proposed changes in flight plans for individual flights in terms of all of the relevant considerations. Many of these are safety checks that are normally handled by the dispatcher, such as considering MELs, and safe landing weights (Smith, McCoy, et al., 1997).

Based on these considerations, it was recommended that, rather than trying to replicate AOC capabilities at NASA, distributed simulation capabilities should be developed which would allow dispatchers participating in a simulation to work at their AOCs with access to all of the tools and information needed to perform dispatch functions. This would also require that, in the design of a simulation, datasets would have to be developed to support schedule management, congestion management and flight planning considerations (passenger connections, crew times, maintenance schedules, MELs, gate availability, NOTAMS, winds and weather, traffic forecasts, ATCSCC advisories, etc.). By doing so, not only are more thorough and realistic simulations likely to be possible, in addition NASA may be able to reduce costs by avoiding the need to develop simulation tools that replicate the functions of existing airline software packages (such as flight planning software that can consider control by time of arrival).

Strategic control in the enroute environment can be divided into two categories, preflight decisions and decisions made while enroute that have a "significant" impact on the time, fuel and/or airspace traversed by a flight, or that have a "significant" impact on traffic flows or on how that flight contributes to the airline’s schedule. As part of this review, we focused primarily on the latter (enroute changes), as that was the area of interest to NASA. However, a few general observations regarding preflight decisions merit emphasis before dealing with enroute changes.

The literature reviewed on AOCs, as well as the structured interviews, indicated five guiding principles regarding design of the future airspace system in terms of preflight decision making. The first is that, consistent with the broadest definition of "free flight" as contained in the original RTCA document (Air Line Pilots Association, 1996; Baiada, 1995; Baiada, 1996; Baker, 1995; Ball, et al., 1995; Blattner, 1997; Corwin, 1997; Cotton, 1995; FAA, 1995; Planzer, et al., 1995; RTCA, 1997; RTCA Task Force 3, 1995a, 1995b, 1995c, 1995d; Wickens, et al., 1997; Wickens, et al., 1998), the system should be designed to provide the users with as much flexibility as possible in determining departure times, routes, altitudes, speeds and arrival times, so that the airlines can make choices based on their business concerns. The second principle is that, in general, increases in system capacity must be a very high priority in designing the future functioning of the airspace system. Designs that provide more flexibility at the expense of significant reductions in capacity are not desirable.

A third principle is that AOCs should be given the best possible information about system constraints (such as weather, traffic and special use airspace availability), and that they should be allowed to develop their own solutions (such as filing routes through Canada to avoid a weather constraint in ZAU) before solutions are imposed. The fourth principle is that, when some type of solution needs to be imposed (such as the specification of a set of alternative reroutes that must be used), decisions made by the ATSP should be made through a collaborative process incorporating input from system users. The fifth guiding principle is that, when necessary, imposed solutions should exert control at the highest level of abstraction possible in order to give the system users as much flexibility as possible.

These principles have several implications. One conclusion is that there is a great need to develop tools that allow AOCs to predict traffic congestion, so that they can make effective decisions about how to plan schedules in terms of the departure times, routes, speeds, altitudes and arrival times for individual flights. As discussed above, tools like FACET offer promise in this area.

A second conclusion is that procedures (and tools to implement these procedures) need to be developed to equitably allocate airspace to the different system users when flights are filed or amended. Because users sometimes compete for the same airspace, there needs to be a mechanism that clearly and easily determines allocations.

There are many alternatives for the design of the enroute environment. These alternatives differ in terms of the degree of structure imposed on traffic flows. At the one extreme is the traffic flow management (TFM) system in place prior to 1995. Generally speaking, at that time the ATSP operated under a control by permission paradigm (Smith, McCoy, Orasanu, et al., 1997), specifying preferred routes and severe weather routes for the users, and requiring them to file and fly these routes unless the ATSP by exception approved a user-requested alternative either pre-flight or while enroute. At the other extreme is the RTCA proposal for enroute "free flight", a control by exception paradigm allowing the users to file and fly whatever routes they desire with minimal imposed structure.

In addition to these two extremes, a number of different approaches to structuring the enroute environment were identified as part of this effort. These alternatives focus on different strategies for structuring the airspace. One example is a proposal to identify 4 dimensional tubes (defined in terms of time and space) that are one-way highways within which flights can pass each other. Such tubes could bend through the airspace in order to avoid constraints, but would impose structure on choices during preflight planning and on strategic ("significant") changes in flight plans while enroute.

Under this proposal, flexibility regarding strategic solutions would be provided to the users in the sense that, for any given city pair, the user could select from a number of different pathways through these tubes. Another aspect of flexibility would be the fact that this network of tubes could also be redesigned on days where weather constraints or traffic demands called for a differently designed network. (Obviously, there are significant issues that need to be considered in terms of the procedures and tools necessary to implement such a system.) Comments by dispatchers regarding this approach are listed below.

A second approach to a structured enroute environment uses the analogy of ribbons in space, rather than tubes:

In addition to these alternatives, there was a recommendation to consider designing additional "reservoirs" within the airspace. The argument for this is that there is always uncertainty associated with plans in the NAS. As a result, if full use of the available capacity is to be achieved, reservoirs must be available to provide extra aircraft to opportunistically take advantage of "slots" that become available. Note that this strategy could be implemented with a variety of different approaches to structuring (or not structuring) the airspace.

This is similar to the practice of using pre-designed holding fixes near major airports to ensure that there are always sufficient aircraft available to fill all of the landing "slots" that become available. These reservoirs would not necessarily all use circular holding as the method for creating a reservoir, however. Another example would be to incorporate in some flight plans the intention of landing short at another airport if the desired destination cannot be reached.

Above, a wide range of alternatives were outlined for strategic control in the enroute environment. These ranged from unstructured strategic control ("enroute free flight") to structured approaches that give the airlines options within dynamically generated traffic flow structures.

The dispatchers interviewed unanimously agreed that unstructured strategic control ("free flight") would not work in congested areas of airspace (such as the Northeastern United States). They indicated that such an approach would not make it possible to safely produce sizeable increases in system capacity.

Some of the dispatchers interviewed did indicate, though, that they thought that in regions without significant air traffic congestion, it might be feasible to allow unstructured strategic control ("free flight") where preflight filings and "significant" flight plan amendments while aircraft are enroute could be made without imposing structure on the overall traffic flows:

"I just don’t think it is practical. At least not into the Northeast. Now if you’re looking at Seattle, Portland, Salt Lake City, probably Denver, places like that with lots of airspace, lots of different directions people are coming into, it might work great. There’s a lot of airspace that’s not used out there. And if you take away the constraints of the VORs and airways so they could fly as they wanted, you would have airspace available to implement a system like that. However, if you start to push into the Northeast, you may have 500 airplanes heading for 8-10 runways depending on wind configurations It becomes strictly a matter of you just can’t squeeze that much in that area in a timeframe like that. … It’s the old analogy of 50 kernels of corn in a barnyard with 500 chickens."

Whether an unstructured or structured approach is used in a specific portion of the NAS, however, the dispatchers reiterated the importance of participation and concurrence when "significant" changes in a flight plan are to be made enroute. These beliefs are summarized in the quote from a dispatcher below.

Several dispatchers further suggested that, even though the approval process for initial flight plans or strategic amendments must involve all three groups (the flight crew, dispatch and the ATC system), during preflight planning it is desirable to give the airlines sufficient information about system constraints so that they can make the initial proposals for routings. This perspective is provided in the following quotes.

"Think of airspace as a commodity. Consider flights from ATL to DFW. The computer can generate a square box around each chunk of airspace where a plane could be. That airspace belongs to the airplane in the box. There’s plenty of space or boxes between ATL and DFW. … When I plan a route, I need to know if all of the boxes that the line [route] goes through are empty. And if they do need miles-in-trail, then the boxes can be spaced further apart. …We could use that information to plan routes. We could also use it for quality control. Every empty box is potentially wasted airspace. We could go back Monday morning and say: What went wrong? How many unfilled boxes were there?"

It is important to note that these statements seem to imply that, if traffic managers looked at the traffic flows after the airlines had filed routes for "avoiding hazards, whether it's traffic or weather," the traffic managers would agree that the resultant flows solved most of the problems of safely avoiding constraints. Some dispatchers, though, made an even stronger statement:

"They [traffic managers] are trying to micromanage everything. It would be better to get them out of the decision making when it’s about weather. … There aren’t that many solid lines of thunderstorms. Let the pilots use their skills to deviate around the storm cells. … When they move those planes away from the weather, they are creating more problems than they are solving."

In short, this suggests that dispatchers and traffic managers may disagree in their judgment of what are "acceptable" traffic flows when there is weather. Technology can help ensure that everyone is considering the same data, but it does not guarantee that they will arrive at the same conclusion.

In some cases, this disagreement is because each party is considering a different set of constraints to select a route. As an illustration, the figure below shows a flight from Kennedy to Portland that was filed basically on a great circle direct (the light blue route). When that airplane was getting ready for takeoff, it was given a reroute that took the airplane from New York to Atlanta, to Houston, to Phoenix, to San Francisco and then to Portland, adding in the neighborhood of four hours to the flying time to this flight (the red route).

As the dispatcher noted:

"The airplane would have had to do aerial refueling to have enough fuel for the

flight. All this was done without any consultation with the dispatcher. The person that came up with this route in the first place had no idea of the fuel capabilities of the airplane."

In other cases, however, there is a fundamental disagreement about the nature of the situation (such as the likelihood that the weather will make a route impassable even with deviations around storm cells, and that this will result in an unacceptable situation in terms of traffic separation).

As illustrated by the enhanced Ground Delay Program (GDP), there are a variety of levels at which the ATSP can exert control if it is needed. In the case of GDPs, the current system has evolved from control in terms of the specific flights allowed to arrive at a given airport in a given time period to control in terms of a limited number of available landing slots per airline at that airport.

As illustrated by proposals to control by time of arrival (CTA), this same possibility arises in terms of strategic control in the enroute system. Under these proposals, a flight can depart whenever the airline desires (subject to any departure airport constraints), as long as its plan ensures that it will get to a specific arrival fix at the arrival airport at a pre-specified time. Such proposals could also incorporate the concept of substitutions of one flight for another to use the slot associated with a particular CTA, with reassessments when the flight is at top-of-climb (thus knowing that it has now departed) and at later points while enroute, as illustrated by the following statement by a dispatcher.

"With the controlled time of arrival strictly standing by itself, I ought to determine what airplane I want to arrive at that time."

Note that such proposals regarding the use of CTAs as the parameter of control used by the ATSP could be implemented under a number of the different strategies discussed above for structuring the airspace system.

Below are representative comments by the interviewed dispatchers who were supportive of this approach.

It should be noted, though, that not all of the dispatchers agreed that use of CTAs as the parameter of control is feasible. As two dispatchers cautioned:

Another pair of dispatchers pointed out other potential complexities associated with coordination if CTAs are used as the parameter of control:

Dispatcher 1: "If they tell me that my flight can get to the approach fix at LaGuardia at 2300 Zulu, and they don’t care when it leaves Minneapolis, well, I may believe, based on my internal models to say, that eight times out of ten, if I get there half-an-hour earlier, they'll have a slot for me without compromising the flow. So I may be willing to take the risk to launch to get there half-hour early with sufficient fuel to hit my control time of arrival and hit my RTA, or to hit my fixed clearance time."

Dispatcher 2: "But there's a problem with this, because you're going to do that, and so am I."

Dispatcher 1: "Right."

Dispatcher 2: "And first thing you know, we're going to have 12 extra airplanes holding at that approach fix and there's no space for them. So how do you deal with that? Do you bring the traffic manager in as part of that decision?"

The above discussions regarding strategic control in the enroute environment emphasize the goal of safely increasing enroute capacity. Another potential benefit of new paradigms for strategic control is the ability for airlines to more flexibly amend flight plans to meet the priorities imposed by their business concerns or to increase throughput by filling in potentially unused slots. These issues were alluded to in the following quotes.

The dispatchers interviewed concurred that a system that provided more flexibility was highly desirable. They also agreed that flight amendments of this type should be viewed as strategic changes that require dispatcher involvement (with appropriate automation support), even when the impact on the performance of the individual flight (in terms of time or fuel) might be small, as by definition such changes involve considerations from a broader perspective (airline scheduling goals and constraints). This view is illustrated by the following quotation.

"What the crew needs is a set of good alternatives that so that they can evaluate

One dispatcher expressed a different viewpoint, though, suggesting improved technologies to deal with minor adjustments of sequences at the arrival fix are not likely to be of practical significance. He suggested that:

"When it’s really important, like if we have a medical to deal with or are short on fuel, we can go to the Center now and ask for priority. The same is true for getting permission to give priority to a flight in a holding stack that is getting low on fuel. We’ve got bigger problems. We can already deal with this when it comes up."

The discussion above indicates that there are a variety of alternatives for developing a more or less structured airspace system. These alternatives vary from minimal structure to a variety of flexible but highly structured solutions (both during preflight planning and for strategic or "significant" amendments while a flight is enroute). The unanimous opinion of the dispatchers interviewed was that, for congested regions of the airspace, there needs to be significant structure in order to achieve high capacity, and that it is not acceptable to provide greater flexibility at the expense of significantly reducing capacity. They also agreed, however, that this structure must be dynamic so that it can be adapted to changing constraints that arise in the airspace, and that within the context of this structure, options should be available to the users. Considerable research is needed to define and evaluate alternative proposals for structuring the airspace.

In addition, given the uncertainties associated with predicting performance in the NAS, it was recommended that methods for incorporating additional reservoirs be evaluated in the design of any airspace structure under consideration. Finally, it is recommend that tools and procedures be studied regarding the implementation of different parameters of control within alternative proposals for structuring the airspace.

The above discussion regarding strategic control in the enroute environment was included in part because the participating subject matter experts unanimously agreed that the viability and cost-effectiveness of alternative approaches to tactical control depend on the specific strategic environment in which they are embedded.

As an example, there was unanimous agreement that self-separation for tactical adjustments (Bowers, 1996; Carlson, et al., 1996; Duong, et al., 1997; Geddes, et al., 1996; Goodchild, et al., 1999; Hoekstra, et al., 1998; Hoekstra, et al., 1999; Hoekstra, et al., 2000; Lozito, 1997; Parasuraman, et al., 2001; Proctor, 1997; Sastry, 1997; Scallen, et al., 1996; Scallen, et al., 1997; Smith and Hancock, 2000; Smith, Woods, et al., 1998; Smith, Woods, et al., 1999; Valenti, et al., 2000; Valot and Grau,1995; Warren, 1997; Wise, 1996a, 1996b) is not a viable solution in an unstructured strategic environment where a high throughput is required, both because of safety concerns and because of a belief that this tactical solution would unacceptably limit capacity or throughput. As an example, one dispatcher commented:

"I’m all in favor of letting the pilot make tactical decisions on minor deviations to avoid weather. For many of these all it takes is a minor deviation of 20 miles around a cell instead of taking a several hundred mile reroute. But for traffic separation, I think a controller is in the best position. If there are 3 or 4 aircraft, it’s got to be the controller doing that."

In contrast, many of the dispatchers interviewed indicated that self-separation might be effective in a structured, high-capacity enroute environment, such as the design based on one-way "tubes" that was discussed above. This judgment is reflected in the following statements regarding self-separation.

"Controllers make deviations in the small area, during approach, departure, even enroute, that we know nothing about today. As long as it's a small deviation -- and we can argue over that definition, I don't think we have any problem with that. But when you start setting up small deviations that lead to large deviations, by the time you go from New York to Seattle, that's a problem. So one, you have to figure out, or be careful of where the small one leads you. There’s a difference between traffic separation, whether it's cockpit based or ATC based, and flight planning which requires joint responsibility for the conduct of the flight."

From a dispatcher’s perspective, by definition new methods for tactical control should not involve changes that have a "significant" impact on the efficiency of a single flight, as by definition such adjustments are small. Furthermore, such tactical amendments should not have an impact on how that flight affects the airline’s schedule. Thus, under this definition of "tactical", the important issue regarding alternative methods of tactical control, such as self-separation, is whether they can safely increase traffic flows (in order to increase throughput or system capacity).

Three general alternatives for tactical control within the enroute environment were identified:

• Self-separation, where flight crews (supported by appropriate technologies) have positive control and make tactical adjustments, with the ATSP playing a monitoring role, intervening only if some need is perceived and positive control is handed off from the flight crew to the ATSP.
• Automated separation, where technologies monitor conditions and make tactical adjustments as necessary (with the flight crew and the ATSP monitoring performance in case human intervention is necessary).
• Enhancements of the current tactical environment, where ATC has positive control (with the concurrence of the flight crew). Such changes could involve changes in the staffing or distribution of work among air traffic controllers, or the introduction of new technologies to enhance controller performance.

From an AOC perspective, a critical question is which of these alternative operational concepts would be most effective in increasing traffic flows and improving safety. A second consideration is the ability of these alternative operational concepts to allow the airlines to prioritize flights in order to avoid diversions or to better achieve other airline scheduling goals. (This second consideration, however, involves strategic rather than tactical amendments to a flight plan, and therefore was discussed earlier in the section on strategic control.)

Thus, from an AOC perspective, the scenarios that need to be evaluated to assess these alternative approaches to tactical control are scenarios assessing the potential to significantly increase traffic flows in order to increase system capacity, and scenarios that critically assess each alternative operational concept in terms of safety. Clearly, however, to test these alternative concepts for tactical control in the enroute environment, the alternative strategic environments in which they will be applied must first be defined. For instance, the effectiveness of a concept like self-separation to increase capacity or improve safety may be quite different in an airspace system with structured flows than in an airspace system without any imposed structure.

In short, in order to evaluate alternative operational concepts for tactical control in the enroute environment from an AOC perspective:

• Alternative operational concepts for strategic control (preflight and enroute) must first be defined.
• Scenarios then need to be developed for each of these alternative operational concepts for strategic control to test the effectiveness of each of the different operational concepts for tactical control in that particular strategic environment. These scenarios need to evaluate whether, in a particular strategic environment, a specific operational concept for tactical control contributes to an increase in system capacity or to an improvement in safety.

Finally, regarding tactical control, the view of the dispatchers interviewed was that the roles of the dispatcher would likely remain similar to current roles regardless of the concept implemented for tactical control:

• Providing look-ahead by monitoring for developing situations where tactical adjustments might be required, alerting the flight crew to these situations ahead of time and indicating possible tactical responses.
• On a limited basis, providing specific recommendations for tactical responses to weather when the dispatcher notes a situation requiring an immediate response, and where the dispatcher may have information or insights not available to other team members (the flight crew and the ATSP).

Regarding the evaluation of alternative methods of tactical control in the enroute environment, such as self-separation, automated separation or enhancements of the current system of positive control by ATC, the above discussion indicates that, from an AOC perspective, the following types of scenarios merit exploration:

• Scenarios Considering Traffic Structuring: The potential effectiveness of a particular method of tactical control such as self-separation depends upon the strategic environment in which it is embedded. Simulations to assess tools and procedures to support self-separation (or some other approach) as a method of tactical control in the enroute environment should therefore consider scenarios based on a range of different methods of structuring the enroute environment (including as an extreme case an unstructured enroute environment). These scenarios should also explore a range of different traffic densities or throughputs.

• Scenarios Considering Strategic vs. Tactical Solutions: Strategic planning,

• Scenarios Focusing on Safety: Within all of the various scenarios suggested inthe preceding two bullets, potential responses to safety critical events need to be considered.

The previous bullets focused on scenarios for studying methods for tactical control in the enroute environment. There are also a number of considerations for developing simulations to study methods for strategic control in the enroute environment. As with the scenarios considering new methods for tactical control, scenarios considering preflight control need to focus on the impact on traffic throughput, predictability and safety. In addition, metrics need to be identified to evaluate the impact of alternative approaches to preflight control on the airlines’ performance in terms of their schedules and business concerns.

Different classes of scenarios for strategic control in the enroute environment are categorized below.

• Scenarios Concerned with the Preflight Locus of Control: One of the hypothesesposited by the dispatchers was that, with access to appropriate information and the tools to process this information, many bottlenecks and inefficiencies could be resolved as a result of improved flight planning and schedule management by AOCs. A variety of simulations could be designed to study variations in the locus of control for preflight planning (AOC vs. TFM), variations in the paradigm for strategic control (control by directive vs. control by permission vs. control by exception vs. control by collaboration), and variations in the design of tools to support information access and decision making by individuals as well as coordination and collaboration among different individuals.

• Scenarios Dealing with the Preflight Parameter of Control: In addition to alternatives for the locus of control, there are alternatives for the parameter of control. To the extent that some organization (such as the ATSP) or tool (such as FACET) places constraints on the options available to AOCs during preflight planning, the impact of the parameter used to exert control needs to be studied. In the current NAS, an example is the use of assigned reroutes to restrict preflight planning by the airlines during weather events. As discussed in earlier sections of this report, the use of CTAs is another example.

• Scenarios Focusing on Preflight and Enroute Constraints Due to the Structuring ofthe Airspace: As with alternative methods of tactical control, the effectiveness of different approaches to preflight planning and strategic adjustments while enroute are likely to be influenced by the degree and nature of structuring of the enroute airspace. A number of different approaches were discussed earlier, ranging from unstructured enroute airspace (enroute "free flight") to flexible but structured alternatives. Note that these variations could also consider different ways to build in reservoirs to help avoid wasted "slots".
• Scenarios Dealing with Flexible Strategic Control while Enroute: The dispatchers all agreed that, when a situation arises while a flight is enroute that requires or allows a "significant" change in its flight plan, dispatcher involvement is necessary to ensure consideration of all of the relevant safety and efficiency issues. (Note that such a situation could be the failure of a predicted event such as a storm to develop as well as the unexpected development of some new constraint.) As with preflight planning, scenarios need to be considered that vary the locus of control, the parameter of control, the timing of the intervention (degree of look-ahead) and the constraints on available alternatives (such as structuring of the airspace).

• Scenarios Concerned with Transitions: Increased capacity within some some enroute portion of the airspace is of limited value if some other constraint is the limiting factor in determining throughput. A number of the potential limiting factors are outside the scope of this report (such as airport runway capacities). However, there is a potentially important interaction between tactical and strategic strategies in the enroute airspace and the capacity to safely handle transitions into terminal area airspace.

The focus of this report has been on AOC perspectives regarding the design of the future NAS, with a particular emphasis on the incorporation of strategic and tactical "free flight" into such a system. A number of directions have been identified for future research and development activities. These directions emphasize the need to take a systems view of performance that considers throughput for the NAS as a whole, schedule management for an airline’s fleet, and flight-specific concerns regarding safety and efficiency. These research directions also emphasize the need to view the NAS as a distributed system with participation by AOCs as well flight crews and the ATSP.

Avans, D. and Smith, K. (1997) Experimental Investigations of Pilot Workload in Free Flight. Proceedings of the Ninth International Symposium on Aviation Psychology, 893-899.

Air Line Pilots Association. (1996). Free flight perspectives on plans, progress and potential from the pilot’s viewpoint. 1996 RTCA Symposium Proceedings, 37-54. Fort Worth, TX.

Baiada M. (1996). Free flight- Rapid implementation. Journal of ATC, July-September. 32-38.

Baiada, M. (1995) Free Flight: Reinventing ATC. Proceedings of the Eighth International Symposium on Aviation Psychology, 410-415.

Baker, W. J. (1995). Free flight: The pilot’s perspective. Journal of ATC, January-March, 46-47.

Ball, M, DeArmon, J. S., & Pyburn, J. O. (1995). Is free flight feasible? Results from initial simulations. Journal of ATC, January-March, 14-17.

Beatty, R. (1995). Airline Operational Control Overview. Airline Dispatchers Federation and Seagull Technology, Inc. Cooperative Research and Development Agreement 93-CRDA-0034, Washington, D.C.

Beatty, R. and Smith, P.J. (2000). Design Recommendations for an Integrated Approach to the Development, Dissemination and Use of Reroute Advisories. Institute for Ergonomics Technical Report #2000-13. Ohio State University, Columbus OH.

Beatty, R., Smith, P.J. and Billings, C. (2001) Design recommendations for an integrated approach to the development, dissemination and use of reroute advisories. Proceedings of the 2001 Aviation Psychology Symposium, Columbus OH.

Billings, C. (1997). Aviation Automation: The Search for a Human-Centered Approach. Mahwah, NJ: Erlbaum.

Blattner, L., (1997). Figuring out free flight. Air Line Pilot, 66. January 1997, 26-29.

Bowers, Kerri L. (1996). Determining the feasibility of a flight profile in a free flight environment. Proceedings of the AIAA/IEEE Digital Avionics Systems Conference, Atlanta, GA.

Braune, R. Jahns, D., Bittner, Jr. , A. C. (1996). Human factors systems engineering as a requirement for the safe and efficient transition to free flight. Proceedings of the Human Factors and Ergonomics Society 40th Annual Meeting, 102-105.

Carlson, L.S., Rhodes, L.R., Cullen, M.G. (1996). Effects of Unstructured Routes on En Route Controllers’ Work Activities and Operational Environment. Mitre/CAASD, McLean, VA.

Corker, K., Gore, B. F., Fleming, K., and Lane, J. (2000). Free Flight and The Context of Control: Experiments and Modeling to Determine the Impact of Distributed Air-Ground Air Traffic Management on Safety and Procedures. 3^rd USA/Europe Air Traffic Management R & D Seminar, Napoli. http://atm-seminar-2000.eurocontrol.fr/acceptedpapers/index.htm

Corwin, B. Honeywell Technology Center. (1997). Free Flight in the Terminal Area: Where Real Capacity Improvements Can Be Made. Proceedings of the Ninth International Symposium on Aviation Psychology, 900-905.

Cotton, W. B. (1995). Free flight in domestic ATM. Journal of ATC, January-March, 10-13.

Duong, Vu N., Hoffman, E., Nicolaon, Jean-Pierre. (1997) Initial Results Of Investigations Into Autonomous Aircraft Concept (Freer-1). Air Traffic Management R&D Seminar, Saclay, France June 17-20. http://atm-seminar-97.eurocontrol.fr

Endsley, M. R. (1997) Situation Awareness, Automation & Free Flight. Air Traffic Management R&D Seminar Saclay, France June 17-20. http://atm-seminar-97.eurocontrol.fr

Endsley, M. R., Mogford, R. H., Allendoerfer, K. D., Snyder, M. D., Stein, E. S. Effect of Free Flight Conditions on Controller Performance, Workload, and Situation Awareness. http://www.cami.jccbi.gov/AAM-500/freeflight/mre_doc.html

FAA Free Flight Website, About Free Flight, http://ffp1.faa.gov/about/about.asp

FAA Free Flight Website, Free Flight Metrics, http://ffp1.faa.gov/approach/approach_ben_met.asp

FAA Free Flight Website, Free Flight Reports to Congress,

http://ffp1.faa.gov/about/about_milestones_congress.asp

FAA Free Flight Website, Expectation Reports on Free Flight Tools, http://ffp1.faa.gov/about/about_ffp1.asp

FAA Free Flight Website, Free Flight Services, http://ffp1.faa.gov/tools/tools.asp

FAA Free Flight Website, Free Flight Stakeholders. http://ffp1.faa.gov/about/about_stakeholders.asp

FAA. (1995). Advancing free flight through human factors: Workshop report. Technical Workshop Report, FAA AAR-100, Washington DC.

Geddes, N.D.,W.S. Beamon, III, & J.A. Wise, J.L. Brown, W. Hamilton, R. Lee, J. Zyzniewski, P. Aumon, & T. Player. (1996). A Shared Model of Intentions for Free Flight. Center for Aviation/Aerospace Research Technical Report No. CAAR-15003-96-01. September. Prepared for NASA Ames Research Center under Contract C-NAS2-14287.

Goodchild, C., Vilaplana, M. A. and Elefante, S. ATM Research Group, Department of Aerospace Engineering, University of Glasgow. (1999). Researchers Study Methods For Ensuring Cooperative Airborne Separation in Free Flight Airspace. ICAO Journal, 54 (9). http://www.zimbo.freeserve.co.uk/CNSATM/Publications/pub2.html

Hilburn, B.G., Pekela, W.D. Parasuraman, R. (1997). The Effect of Free Flight On Air Traffic Controller Mental Workload, Monitoring and System Performance. http://www.cami.jccbi.gov/AAM-500/freeflight/research.html

Hilburn, B.G., Free Flight and Air Traffic Controller Mental Workload. Aviation Psychology Symposium, Columbus, Ohio, 1997, http://www.nlr.nl/public/hosted-sites/freeflight/ffdoc.htm

Hilburn, B.G., Bakker, W.P.M. and Pekela, W.D. (1997). The Effect of Free Flight on Air Traffic Controller Mental Workload, Monitoring and System Performance. CEAS Symposium on Free Flight, Amsterdam. http://www.nlr.nl/public/hosted-sites/freeflight/ffdoc.htmhttp://www.nlr.nl/public/hosted-sites/freeflight/ffdoc.htm

Hilburn, B.G., Pekela, W. D. Air Traffic Controller Strategies in Resolving Free Flight Traffic Conflicts: The Effect of Enhanced Controller Displays for Situation Awareness. World Aviation Congress, Anaheim, 1998, http://www.nlr.nl/public/hosted-sites/freeflight/ffdoc.htmhttp://www.nlr.nl/public/hosted-sites/freeflight/ffdoc.htm

Hoekstra, J.M. Ruigrok, R.C.J. and Van Gent, R.N.H.W. (2000). Free Flight in a Crowded Airspace. ATM 2000 Conference, Napoli, June 2000. http://www.nlr.nl/public/hosted-sites/freeflight/ffdoc.htmhttp://www.nlr.nl/public/hosted-sites/freeflight/ffdoc.htm

Hoekstra, J.M. Ruigrok, R.C.J., Van Gent, R.N.H.W. (1999). Designing for Safety: the Free Flight ATM Concept. Human Error, Safety & System Design Workshop, Liege, Belgium. http://www.nlr.nl/public/hosted-sites/freeflight/ffdoc.htm

Hopkin, V.D. (1995). Human Factors in Air Traffic Control. New York: Taylor-Francis.

Kerns, K, Smith, P.J., McCoy, C.E. and Orasanu, J. (1999). Ergonomic issues in air traffic management. In W. Marras and W. Karwowski (eds.). Handbook of Industrial Ergonomics. CRC Press, 1979-2003.

Krozel, J. and Mogford, R. (September, 2000). Free Flight Research Issues and Literature Search. Los Gatos, CA: Seagull Technology, Inc.

Lacher, A. and Klein, G. (1993). Air Carrier Operations and Collaborative Decision-Making Study, MTR 93W0000244, The MITRE Corporation, McLean VA.

Lozito, S., McGann, A., Mackintosh, M. A. and Cashion, P. (1997). Free Flight and Self-Separation From The Flight Deck Perspective. Air Traffic Management R&D Seminar, Saclay, France June 17-20. http://atm-seminar-97.eurocontrol.fr

Mitre/CAASD, Website. Collaborative Routing Coordination Tools. http://www.caasd.org/proj/crct/

Nordwall, B. (1995). Cultural shift key to concept. Aviation Week & Space Technology, July 31, 1995, 40-41.

Orasanu, J. (1991). Shared mental models and flight crew performance. In Johnson, McDonald and Fuller (eds.), Aviation Psychology in Practice. Brookfield, VT: Avebury, 255-285.

Orasanu, J. and Salas, E. (1993). Team decision-making in complex environments. In G.A. Klein, J. Orasanu, R. Calderwood and C.E. Zsambok (eds.), Decision Making in Action: Models and Methods. Norwood, NJ: Ablex.

Parasuraman, R., Duley, J., Masalonis, A., Galster, S., Castano, D., Metzger, U., Wang and F., Hilburn, B. G. (2001). Human Factors in Free Flight: Developing Dynamic Automation Tools to Support Air Traffic Management. http://www.cami.jccbi.gov/AAM-500/freeflight/research.html

Prinzo, O.V. (2001). Civil Aeromedical Institute. Pilot Visual Acquisition of Traffic: Operational Communications From OpEval-1. Final Report.

Planzer, N. and Jenny, M. T. (1995). Managing the evolution to free flight. Journal of ATC, January-March, 10-13.

Proctor, P. (1997) Updated ATC software presages free flight. Aviation Week and Space Technology, January 27.

Rasmussen, J., Brehmen, B. and Leplat, J. (eds.) (1991). Distributed Decision Making: Cognitive Models for Cooperative Work. Chichester, U.K.: Wiley.

Robertson, S., Zachery, W. and Black, J. (eds.) (1990). Cognition, Computing and Cooperation. Norwood, NJ: Ablex.

RTCA. (1997). Operational Concepts and Data Elements Required to Improve Air Traffic Management (ATM) – Aeronautical Operational Control (AOC) Ground-Ground Information Exchange to Facilitate Collaborative Decision Making. Document No. RTCA/DO-241.

RTCA Task Force 3 (1995a) Free Flight in the Real World. 1995 RTCA Symposium Proceedings, 13-91. Seattle, Washington.

RTCA Task Force 3 (1995b). Executive summary of the Final Report of RTCA Task Force 3: Free flight implementation, RTCA, Inc., Washington DC.

RTCA Task Force 3 (1995c). Final report of RTCA Task Force 3: Free flight implementation, RTCA, Inc., Washington DC

RTCA Task Force 3 (1995d). Interim Report on free flight implementation. RTCA Paper No. 506-95/t3-34, RTCA, Inc., Washington, DC.

SAE G-10W Free Flight Subcommittee: Human Factors Issues in Free Flight. SAE G-10 Aerospace Resource Document (ARD) # 50079 http://www.cami.jccbi.gov/AAM-500/freeflight/ard.html

Sastry, S. (1997) Hybrid Control Issues In Air Traffic Management Systems. Air Traffic Management R&D Seminar, Saclay, France June 17-20. http://atm-seminar-97.eurocontrol.fr

Scallen, S. F., Smith, K., and Hancock, P. A., (1996). Pilot actions during traffic situations in a free flight airspace structure. Proceedings of the Human Factors and Ergonomics Society 40th Annual Meeting, 111-115.

Scallen, S. F., Smith, K. and Hancock, P.A. (1997). Influence of Color Cockpit Displays of Traffic Information on Pilot Decision Making in Free Flight. Proceedings of the Ninth International Symposium on Aviation Psychology, 368-373.

Sheridan, T. (1987). Supervisory control. In G. Salvendy (ed.), Handbook of Human Factors. New York: Wiley.

Sheridan, T. (1992). Telerobotics, Automation and Human Supervisory Control. Cambridge, MA: MIT Press.

Smith, K. and Hancock, P.A. (2000). Traffic maneuverability and cockpit display characteristics determine whether commercial airline pilots can maintain self separation in realistic scenarios of en-route flight. 10th European Conference on Cognitive Ergonomics (ECCE-10), Linköping, Sweden, August 21-23.

Smith, P.J., Billings, C., Woods, D., McCoy, E. (1998). Human factors issues in a future air traffic management system. Proceedings of the 17th Digital Avionics Systems Conference, 1-7.

Smith, P.J., Billings, C., Chapman, R.J., Obradovich, J., McCoy, E. and Orasanu, J. (2000a). Alternative architectures for distributed cooperative problem solving in the national airspace system. Proceedings of the 5th International Conference on Human Interaction with Complex Systems, Urbana, IL, 203-207.

Smith, P.J., Billings, C., Chapman, R.J., Obradovich, J., McCoy, E. and Orasanu, J. (2000b). Alternative Architectures for Distributed Work in the National Airspace System. Institute for Ergonomics Technical Report #2000-8. Ohio State University, Columbus OH.

Smith, P.J., Billings, C., Chapman, R. J., Obradovich, J., McCoy, E. and Orasanu, J. (2000c). Alternative "Rules of the Game" for the National Airspace System. Proceedings of the Fourth Conference on Naturalistic Decision Making. Stockholm, Sweden.

Smith, P.J., Billings, C., Woods, D., McCoy, C.E., Sarter, N., Denning, R. and Dekker, S. (1997). Can automation enable a cooperative future ATM system? Proceedings of the 1997 Aviation Psychology Symposium, 1481-1485.

Smith, P.J., McCoy, C.E., Denning, R. and Obradovich, J. Heintz. (1997). Cooperative problem-solving in the air traffic management system. Proceedings of the 1997 Annual Meeting of the IEEE Society for Systems, Man and Cybernetics. October 12-15, Orlando, FL, 3137-3141.

Smith, P.J., McCoy, E. and Layton, C. (1997). Brittleness in the design of cooperative problem-solving systems: The effects on user performance. IEEE Transactions on Systems, Man and Cybernetics, 27, 360-371.

Smith, P.J., McCoy, E. and Billings, C. (2000). Issues in the Use of Coded Departure Routes. Institute for Ergonomics Technical Report #2000-12. Ohio State University, Columbus OH.

Smith, P.J., McCoy, E. and Orasanu, J. (2001). Distributed cooperative problem solving in the air traffic management system. In E. Salas and G. Klein (eds.). Linking Expertise and Naturalistic Decision Making. Mahwah, NJ: Lawrence Erlbaum, Associates.

Smith, P.J., McCoy, C.E., Orasanu, J., Billings, C., Denning, R., Rodvold, M., Gee, T. and Van Horn, A. (1997). Control by permission: A case study of cooperative problem-solving in the interactions of airline dispatchers and ATCSCC. Air Traffic Control Quarterly, 4, 229-247.

Smith, P.J., Woods, D., Billings, C., Denning, R., Dekker, S., McCoy, E. and Sarter, N. (1999). Conclusions from the application of a methodology to evaluate future air traffic management system designs. In M. Scerbo and M. Mouloua (eds.), Automation Technology and Human Performance: Current Research and Trends. Mahwah NJ: Erlbaum and Associates, Publishers, 81-85.

Smith, P.J., Woods, D., McCoy, C.E., Billings, C., Sarter, N., Denning, R. and Dekker, S. (1998). Using forecasts of future incidents to evaluate future ATM system designs. Air Traffic Control Quarterly, 6, 71-86.

Transportation Research Circular E-C029, (2001). Aviation Gridlock: Understanding the Options and Seeking Solutions. http://trb.org/trb/publications/circulars/ec029.pdf

Valenti Clari, M.S.V., Ruigrok^, R.C.J. and Hoekstra, J.M. Cost Benefit Study of Free Flight with Airborne Separation Assurance, (2000). AIAA 2000 Guidance Navigation & Control conference, Denver, August. http://www.nlr.nl/public/hosted-sites/freeflight/ffdoc.htm

Valot, C. and Grau, J.Y. Institut de Medecine Aerospatiale du Service de Sante des Armees. (1995) Sharing of Competences Between Pilot and Systems: How to Balance? Proceedings of the Eighth International Symposium on Aviation Psychology, 191-196.

Van Gent, R.N.H.W., Hoekstra, J.M. and Ruigrok, R.C.J. Free Flight with Airborne Separation Assurance http://www.cami.jccbi.gov/AAM-500/freeflight/research.html

Warren, A. (1997) A Methodology And Initial Results Specifying Requirements For Free Flight Transitions. Air Traffic Management R&D Seminar, Saclay, France June 17-20. http://atm-seminar-97.eurocontrol.fr

Weitzman, D. O. (1997). Air Traffic Control Automation: Lessons From Chernobyl And Three Mile Island. Journal of ATC, January to March 1997, 33-35.

Wickens, C. D., Mavor, A. S., Parasuraman, R., and McGee, J. P. (eds.) (1998). The Future of Air Traffic Control: Human Operators and Automation, Washington, D.C.: National Academy Press. http://books.nap.edu/books/0309064120/html/index.html

Wickens, C., Mavor, A. and McGee, J. (1997). Flight to the Future: Human Factors in Air Traffic Control. Washington, D.C.: National Academy Press.

Wise, J.A. (1996). Ergonomics and the problems of cockpit based air traffic systems. Aviation Medicine, Psychology, and Ergonomics, 2, 33-35.

Wise, J.A. (1996). Ergonomics and cockpit centered air traffic systems. Proceedings of the First International Conference on Applied Ergonomics (ICAE ‘96). Istanbul, Turkey. 21-24 May.