ATM Modernization, Business & GA, Commercial

See and Avoid in IMC

By Victor Riley | July 1, 2006
Send Feedback

See and avoid is the primary rule of operation under visual flight rules (VFR). But new flight deck technologies, along with increasing air traffic management (ATM) automation, may someday make see and avoid the norm under instrument flight rules (IFR), as well.

In any airspace, pilots must avoid adverse weather conditions, restricted airspace and other aircraft. They also must be coordinated with other aircraft to optimize the use of airspace during critical phases of flight. Under VFR all of this is done with the visual sighting of other aircraft and radio communications with air traffic control (ATC) to establish who will do what and when. Today, of course, operating under IFR requires direction from controllers. However, FAA recognizes that this may have to change if airspace capacity is to increase enough to meet projected future needs. One vision of this change looks a lot like today’s VFR, where pilots, themselves, will be responsible for avoiding hazards and coordinating with other aircraft. Since the IFR operating environment is much more complex, however, the "rules of the road" also will be more complex, and new onboard systems will be needed.


In 1995 RTCA published a vision of "free flight" involving essentially autonomous aircraft that could avoid hazards and self-separate from other aircraft while establishing and following optimal routes. These routes would no longer be limited to today’s established airways. Aircraft would be able to follow more direct routes based on their individual optimization requirements and current wind, weather and traffic conditions.

Although the phrase "free flight" is heard less frequently today than in the past, the problem of traffic growth and the limited number of controllers hasn’t changed since 1995. Nor have the potential solutions of increased levels of automation and controller-pilot collaboration. Meanwhile, some of the technologies spawned by the free flight concept are just now maturing. This presents FAA with an interesting dilemma: as technology enables more flight deck autonomy, policy makers will have to decide which ATM model, or combination of models, and which mix of technologies to use. To support this effort, the Next Generation Air Transportation System (NGATS) Joint Planning and Development Office (JPDO) has been formed. This multi-agency organization will establish the implementation plan for moving to tomorrow’s more flexible environment.

NASA, in particular, has been actively supporting the free flight vision. Three NASA technologies, when fully mature and combined, may help to make the autonomous airplane possible:

Autonomous Flight Rules

The first piece of this puzzle is a set of autonomous flight rules, developed and prototyped at NASA Langley Research Center (see October 2004, page 32). AFR attempts to resolve basic problems underlying the proposed move to free flight: capacity restrictions arising from controller workload limitations and the consequent underutilization of available airspace.

To recap, the incentive for an operator to adopt and use autonomous aircraft technologies is the ability to operate more flexibly, in a manner less constrained by the conventions of today’s ATC system, in order to control costs. This could include selecting departure and arrival times and choosing routes. Autonomous aircraft, for example, would be able to depart when IFR aircraft are in a ground hold because the former would be able to make a tight arrival time over a metering fix on arrival. In return for greater flexibility and cost control, the more autonomous airplane would have to accommodate the current limitations of unequipped aircraft and ATC requirements.

Since the ultimate objective of the new operational concept is to reduce the capacity constraints imposed by controller limitations, autonomous aircraft must not increase controller workload. They would accomplish this, not only by taking on responsibility for self-separation from hazards, but also by yielding to conventionally controlled aircraft. This would allow the controller to concentrate only on those aircraft that must be positively controlled. The autonomous aircraft would stay out of the way and arrive accurately at their scheduled times.

This concept could increase capacity by shifting responsibility for conflict management from a single controller to multiple pilots, assisted by automation. Since pilots would only be concerned with potential conflicts involving their own aircraft, they need only attend to a small number of other aircraft, while the controller otherwise has to attend to a large number. Each potential conflict generally would have the attention of a pair of pilots, one in each aircraft. Breaking up and distributing the task of separation into many small jobs instead of one large job would allow increased numbers of aircraft to use an airspace without raising controller workload.

Furthermore, since more aircraft would bring more pilots and processors to the task, the AFR concept would scale with the number of aircraft. NASA estimates the concept could improve capacity by as much as 300 percent in some cases, without requiring the controller to serve as a safety backup. In fact, requiring a controller to intervene in a developing conflict situation that the controller has not been involved in would be impractical, says David Wing, a researcher in the Aviation Operations and Evaluation Branch at NASA Langley. The burden would be on the pilots and onboard systems to resolve even the most complex situations without relying on the controller as a backup.

The increased route flexibility that autonomy provides would make it easier for pilots to self-separate. For example, in instrument meteorological conditions (IMC), autonomous aircraft may be able to select routes through the affected airspace, based on their own optimal trajectories, weather observations and weather uplink tools. Furthermore, tools that provide information about other aircraft positions and plans would allow them to coordinate more effectively. One important assumption of the NASA concept is that ADS-B information available to the pilots would be expanded beyond the near-term trajectory projections that the system currently provides to include trajectory intent information derived from the flight management computer’s flight plan and current flight control mode. For example, if the airplane were operating in the LNAV/VNAV mode, the system would broadcast the projected 4D path, enabling better strategic coordination between aircraft.

The AFR concept includes a strategic conflict management system, the Autonomous Operations Planner (AOP), which automatically compares candidate resolution trajectories against all known constraints, including the trajectories of all proximate aircraft. Only those solutions that are completely conflict-free and flyable within the aircraft’s performance envelope are presented to the flight crew for consideration. The pilot can both choose the maneuver and establish some higher-level objectives for the automation, such as minimizing time or fuel impact. However, there may be situations where pilots would want to be more involved in the process of analyzing a situation, exploring the impacts of various options and modifying the route themselves. This is where a cockpit display of traffic information with manual route planning tools can help.

CSD Conflict Probe

Such a tool, the cockpit situational display (CSD) conflict probe, has been under development at NASA Ames Research Center. Its major function is to enable pilots to resolve air-to-air trajectory conflicts. The display shows all aircraft as chevrons pointed in the direction of flight. Aircraft at the same altitude as the pilot’s aircraft are shown in white, with those at higher altitudes in blue and those at lower altitudes in green. Aircraft also can be shown with leading "tails"–lines projecting ahead of the aircraft–depicting their immediate future trajectories, based on ADS-B data. This technique allows pilots to visualize how aircraft will converge. The CSD also has terrain and weather visualization capabilities.

The pilot can gain additional information about any aircraft on the display by placing the cursor over the aircraft icon and selecting it. At the most basic level, the aircraft data block with flight number and altitude can be displayed. The pilot, however, also can bring up a display of the current flight plan with projected turns and altitude changes. This is particularly important for understanding and resolving conflicts involving climbing and descending aircraft. Lastly, the pilot can activate a predictor display that sends a visual pulse along each aircraft’s future trajectory. The combinations of these pulses, which look like small, bright balls originating from the aircraft’s nose and traveling along its path ahead of the aircraft, allow pilots to determine whether their aircraft will cross another aircraft’s path in front of or behind it. If an aircraft will cross another’s path, the pilot can use this predictor display to see which aircraft will reach the crossing position first, based on which aircraft’s pulse predictor reaches that position first.

All of these features help the pilot visualize the surrounding airspace and the nature of potential conflict geometries. However, the pilot is not expected to detect and resolve conflicts without help. The tool also incorporates an automated conflict probe that looks up to 20 minutes ahead of the aircraft along its planned path in order to identify any conflicting trajectories. At cruise altitudes that could translate to about 200 miles ahead. If it finds a potential conflict, it highlights both own aircraft and the conflicting aircraft in yellow. If the alert is merely advisory, such as another aircraft crossing above or below the pilot’s aircraft, both affected aircraft are depicted with yellow outlines and no aural alert is given. However, if the conflict would violate protection zones, an alert is sounded and the affected aircraft are shown in solid yellow. The trajectories of the affected aircraft also are drawn in yellow, so the pilot can see where the conflict is expected to occur.

Once a conflict is detected, the tool’s route modification feature allows the pilot to graphically adjust the route and assess the impact of the change on the projected conflict. Since the conflict analyzer operates in real time, the pilot immediately sees the effects of these modifications. Thus the pilot can quickly determine whether the change would resolve the projected conflict or even create any new conflicts. The pilot can adjust the route by "grabbing" it with a cursor control device and moving it laterally, inserting a waypoint and adjusting the waypoint position or inserting an altitude change segment into the trajectory. Future versions of the tool will also allow speed changes.

The ability to automatically identify conflicts and assess the effects of route changes in real time is particularly useful in cases where conflicts may be difficult to detect or resolve. Shallow-angle conflicts, for example, have been found to be among the most difficult conflict geometries for a pilot to detect and resolve without help, says Walter Johnson, a research psychologist at NASA Ames. The angle of intercept is very small and the conflict point is very far ahead of the aircraft. Conflicts involving multiple aircraft are equally difficult to resolve without visualization tools, particularly in cases where resolving one conflict creates another. Studies demonstrate that pilots using the CSD can detect and resolve conflicts more accurately than without such assistance, and that pilots, when not using the tool, often believe they have resolved a conflict without realizing that they have created another.


Both the Autonomous Operations Planner and the CSD recognize that weather conditions may impose severe constraints on a flight crew’s options. One assumption underlying the autonomous flight rules concept is that pilots may have better weather information than the controllers do and would therefore be better situated to find a route around weather. Potentially the final piece of the autonomous flight deck puzzle, the Advanced Weather INformation system (AWIN) combines a route optimizer with weather and restricted airspace information to find an optimal, hazard-free route. Developed by NASA Langley and Honeywell, AWIN uses uplinked weather information to determine no-fly areas based on convective activity, turbulence, volcanic ash, icing and high ozone levels. These regions are represented as polygons on a graphic display with minimum and maximum altitudes. Restricted airspaces also are represented as no-fly regions.

These no-fly regions are incorporated into the constraints used by the route optimization algorithm to find the best route. Like the flight management system (FMS) cost index, the optimization algorithm represents the pilot’s chosen weighting between saving time and fuel. For example, if the pilot chooses to minimize time without regard to fuel use, the algorithm will find the fastest route. Unlike the FMS, however, which only optimizes the vertical trajectory, AWIN also optimizes the lateral trajectory and avoids the areas designated as no-fly regions, based on atmospheric conditions. This produces the best route for the current conditions, constraints and pilot preferences.

The graphical user interface indicates the type of hazard by the color of the associated polygon and the severity level of the hazard on a numeric scale. Recognizing that the pilot may have other information or preferences, however, designers also have provided slide controls, so the pilot can adjust the acceptable severity levels for each hazard and see what the resulting route would look like. This would allow a cargo carrier, for example, to fly closer to convective activity or accept a higher level of turbulence than a passenger carrier would. The interface also allows the pilot to see only the hazards of interest and depicts them on both a lateral map and on the vertical profile display. This last interface is particularly important for hazards that are altitude-dependent, such as turbulence.

AWIN is still a research prototype and has been evaluated as a dispatch aid in initial tests at Embry-Riddle Aeronautical University. However, Honeywell researchers, Michael Dorneich and Steve Pratt, say that the tool has always been intended ultimately for the flight deck. In a flight deck implementation, the route planning capabilities of the tool could be integrated with the FMS, and the hazard display and pilot interface to the planning capabilities could be hosted on a multifunction display, with hazardous regions shown on the navigation and vertical profile displays. The current impediment to flight deck implementation is the absence of appropriate hazard data standards that would enable the transmission of hazard data to the cockpit. But if the appropriate data services are put in place, the tool could provide pilots with a powerful means of finding hazard-free, cost-optimal routes, while reducing their dependence on ground-based controllers and dispatch services.

Free Flight Visions

The promise of free flight described by the original RTCA report is being realized in the development of enabling technologies and operational concepts. However, the eventual uses and benefits of the technologies may differ somewhat from the original vision. George Donohue, formerly the associate administrator of research and acquisitions for FAA and currently a professor at George Mason University, says that the original concept of free flight was flawed in that it attempted to optimize the en route environment while the real system constraints exist in the terminal areas. Donohue notes that satellite-based navigation and ADS-B provide more precise and timely information than ATC radar, and that, because of the lag time introduced by ground-based control, pilots and avionics are better positioned to take advantage of these improvements for precise control.

Using velocity vector guidance, pilots can self-position their aircraft in an arrival stream much more accurately than the controller can. For example, in a NASA concept called airborne precision spacing, the controller would provide each aircraft with a lead aircraft and a desired spacing at the runway threshold; that aircraft would simply follow automation speed commands to achieve the controller’s desired sequence and final approach spacing.

Flight tested by NASA at Chicago O’Hare in September 2002, this capability allows aircraft to self-adjust their positions in the stream to close unnecessary spaces, or to make room for other arrivals or departures. Donohue cites NASA studies showing that spacing variance can be reduced from around 30 seconds to around 5 seconds, using ADS-B and self-positioning.

One thread that runs through many of the current ATM models is the movement from distance-based control to time-based control. As Donohue notes, this can only be achieved by moving control from the ground to the air. The technologies described here can, in combination, provide pilots with the tools to take on that responsibility.

There is no single path to move these technologies into widespread use in the National Airspace System. There are several concepts of operation for deploying them, including segregating autonomous from controlled aircraft, requiring ADS-B equipage (see story, page 40), and requiring an enhancement to ADS that broadcasts more flight plan-based intent and mode status. Likewise, there are several possible concepts for the role of ATC, ranging from requiring autonomous aircraft to clear proposed trajectory changes with ATC before executing them to requiring ATC to focus only on controlled aircraft and rely on equipped aircraft to take care of themselves. The benefits and drawbacks of these operational concepts are being explored.

Many other organizations are developing similar tools. Technologies such as synthetic vision, for example, can improve aircraft autonomy near the ground. However, the three elements described here are representative of the technologies that, when combined, could help to bring the flexibility of VFR to instrument flight environments. "The FAA is interested in considering the potential implementation of such technologies as part of the total future air transportation system," sums up Kelli Willshire, FAA R&D Field Office manager at NASA Langley.

Receive the latest avionics news right to your inbox