Business & GA, Commercial, Military

Autonomous Flight Management: A Third Way

By David Jensen | October 1, 2004
Send Feedback

Pilots who now can choose to fly under visual or instrument flight rules (VFR or IFR) may someday have a third option: autonomous flight rules, or AFR. Using flight plans and aircraft performance data stored in the flight management system (FMS), along with onboard traffic information, crewmen in AFR-equipped aircraft could make their own decisions -- tactically and strategically -- when changing routes to avoid traffic, restricted airspace or weather. Such freedom from controller direction also requires GPS and a transponder, as well as a decision-aiding tool for flight crews that NASA developed, called the autonomous operations planner, or AOP.

The AOP and accompanying display are designed to permit autonomous flight management (AFM). Two NASA research centers -- Langley in Virginia and Ames in California -- are jointly evaluating the system in bi-coastal, air/ground simulations. Air transport pilots participating in the test program view computer screens in Langley, which show traffic monitored by controllers at Ames. Ames also provides pilots to the simulation.

AFM is part of a much larger NASA endeavor. It is the airborne component of NASA's Distributed Air/Ground Traffic Management (DAG-TM) operations concept, which represents about 20 percent of the research and development agency's Advanced Air Transportation Technologies (AATT) project, launched in 1997.

AATT aims to modernize the U.S. National Airspace System (NAS) by enabling a substantial increase in capacity while maintaining safety. NASA is investigating the DAG-TM concept to determine whether it would be beneficial to distribute decision making for traffic management among flight crews, dispatchers and controllers. AFM is part of the En Route Free Maneuvering element, which explores whether aircraft with advanced equipment can safely manage flight path changes without obtaining approval from the air traffic controller, as required by today's system.

Against the backdrop of the AATT project, which seeks to improve commercial operations all the way from gate to gate, AFM might look relatively minor. But AFM's intent of modifying some of the decision making process would have the following far-reaching benefits:

  • Allowing pilots the freedom to plot the most efficient flight paths, thus reducing airline fuel costs;

  • Allowing pilots to autonomously maneuver safely around traffic and hazards, such as weather, while meeting traffic flow constraints issued by the controllers;

  • Giving dispatchers the flexibility to adjust individual or fleet-wide priorities while aircraft are en route;

  • Reducing controller workload associated with traffic growth; and

  • Increasing airspace capacity -- NASA's "major objective" with the concept, according to Mark Ballin, crew systems branch, research and technology directorate, at NASA Langley.

"We're trying to get around the situation in which the controller can take on no more workload," Ballin explains. That means avoiding having a sector "slam the door" -- the point at which controllers of an en-route sector simply cannot handle greater traffic volume and therefore stop accepting handoffs from upstream sectors.

"We hope AFM will allow an increase in air traffic volume without overloading the sector controller or traffic flow manager with work," says Ballin. "With more information and the AOP, the flight crew adds no more burden to the controller, so now instead of managing the whole sector, he can focus on the aircraft flying IFR." The controller remains in the loop, however; in addition to establishing the traffic flow constraints, he is kept apprised of AFR operations in his sector.

Such a capability could be timely, as U.S. airspace -- and, indeed, airspace worldwide -- is "seeing high traffic demand again since 9/11," according to Ballin. Department of Transportation statistics reveal a steady escalation of air traffic in the United States. NASA estimates that the widespread use of autonomous flight operations could triple airspace capacity.

What also makes AFM appealing is the fact that not all aircraft need to be operating under AFR in order for the concept to benefit both controllers and operators. Therefore, AFM "would be a choice, not a mandate, and it would be up to the carrier to determine whether it is cost-effective," says David Wing, aeronautical engineer, crew systems, at NASA Langley.

Those carriers flying AFR-equipped aircraft will reap the system's benefits, while those that don't will simply proceed under traditional controller direction. The pilots flying AFR will not have to file elaborate fight plans nor follow predetermined waypoints and preplanned altitudes, routes and speeds. With the AOP, all of these variables can be calculated en-route to achieve maximum fuel efficiency, passenger comfort and solve problems. The pilots' only concern will be complying with the traffic flow constraints determined by the controllers. One constraint NASA expects to be commonplace is an assigned arrival time to a terminal arrival fix. How pilots reach that fix is up to them or their dispatchers.

Wing offers a scenario in which the dispatcher -- who also receives traffic, weather and traffic flow information -- might apply AFM: "If an airline has multiple aircraft entering an airport, he may decide to manage the arrivals to the operator's advantage, placing one aircraft [with, say, more passengers making connections] ahead of another."

Current AFR research has focused on en-route and initial descent into capacity-limited terminal airspace. After an aircraft reaches the terminal arrival fix, its crew reverts to IFR, while the controller takes over aircraft separation and manages arrivals and landings in busy terminal airspace.

During the NASA simulations the terminal fix assignments and acknowledgments are exchanged using controller pilot data link communications (CPDLC). The Ames controllers during the simulations apply arrival flow metering to direct the AFR pilots into the terminal area. Used at a handful of the major hub airports, arrival flow metering employs computerized calculations to direct aircraft into individual "slots," rather than having them follow in trail. It is more efficient than the "miles in trail" approach to flow management, in which all aircraft on approach must match the speed of the slowest aircraft, says Ballin. He adds, however, that the AFM concept also can work with miles-in-trail flow management.

The NASA simulations, incidentally, use traffic scenarios taken from the busy Dallas/Fort Worth air route traffic control center. Scripted, high-traffic scenarios are chosen to rigorously test the AFM concept. At Langley, the participating pilots gather in one room, each operating desktop flight simulators showing simulated navigation and flight displays. In an adjoining room Langley technicians view a giant, multi-image screen that shows the traffic flow and pilots' use of the AOP. The Ames laboratory is similarly laid out, with both pilots and controllers operating the simulation.


AFR aircraft will have to be equipped with:

  • A cockpit display of traffic information;

  • Computerized, data link capability to give pilots real-time traffic, flow management and airspace hazard data; and

  • A software system that combines automatic dependent surveillance-broadcast (ADS-B) and traffic information service- broadcast (TIS-B) information with flight plan and aircraft performance data from the FMS via an ARINC 702A bus. It also integrates data from flight information system-broadcast (FIS-B) -- for weather data, notices to airmen, etc. -- via the data link.

To ensure that the aircraft's trajectory (the route, along with speeds and altitudes) planned by the crew is clear of traffic, weather and restricted airspace, the AOP provides multiple maneuvering options. "AOP can present optimized solutions based on various factors selected by the crew -- for example, best minimum fuel or best ride or best time to destination," says Ballin.

Raytheon is the integrator for the AOP program, and a combined NASA/Lockheed Martin team developed the software for the FMS used for simulation. Subcontractors to Raytheon include Titan Systems, which designed the AOP software, following a NASA functional design and using C++ coding. Also on the project are Seagull Technology, developing the traffic simulation, and SAIC, which developed the air crew simulation, called Aircraft Simulation for Traffic Operations Research, or ASTOR. The National Institute of Aerospace (NIA) is furnishing consultants (retired and active airline pilots) to participate in the simulations and "help determine if the AOP functions and AFM procedures are realistic," says Wing.

A production AOP probably would be incorporated in the FMS, but it would have to be in an enhanced unit, as no flight management system available today has the memory to accommodate the software, according to Rick Shay, a commercial pilot and NIA aviation consultant. He adds, however, that the prototype AOP system developed for NASA testing "may well be more complex than a production version."

"We don't know what functions finally will be needed," adds Wing. "That's a long-term research objective."

NASA engineers also believe that, with a growing number of data sources entering the cockpit, a future AOP will incorporate data fusion and "some method of determining the best source of data among multiple sources," according to Ballin. "During our simulations we use one source for one type of data, but that will change."

Meantime, the AOP that NASA is using was modeled after the ARINC 429 bus, to integrate into today's aircraft architecture. It also conforms to current ADS-B performance--but that, too, could change. NASA engineers are participating in a working group within RTCA special committee (SC)-186, tasked to determine minimum requirements for ADS-B.

Enhancing ADS-B?

"Our research may help determine design requirements for ADS-B infrastructure," says Wing. "For example, the transponder might have to deliver more information in terms of aircraft intent." This would mean not just transmitting traffic positioning data, but also flight plan data from the FMS.

Because of bandwidth limits, the current ADS-B system can broadcast only one trajectory change point, such as the top-of-the descent point, waypoint, etc. Up to four trajectory change points are being indicated during NASA simulations, showing more fully an aircraft's intent. "Four would be ideal," says Wing. "But we know that we'll probably eventually have to work with fewer change points, maybe one or two."

During its simulations NASA compared autonomous flight with and without traffic intent information. "We've determined that either is feasible," says Wing. "The intent information clearly is preferred, but without it, pilots still were able to maintain separation."

AOP information would be shown on the pilot's navigation display. A function, called "maneuver restriction bands," on the display helps to prevent traffic conflict. One band, a highlighted arc on the compass rose, indicates the boundaries within which the pilot can maneuver the aircraft laterally, while bands on the vertical speed indicator show the range of vertical speeds and altitude changes to avoid conflict. The Netherlands' national aerospace laboratory (NLR) developed this conflict-prevention concept under NASA sponsorship.

Using the mode control panel and FMS, the pilot can also probe maneuvers and route changes. Some may view the AOP as an airborne version of the conflict probe used by controllers to project aircraft trajectories. But Wing contends that the planner "does more than probing for conflicts. It helps the pilots manage the flight path."

The AOP does much more than the traffic alert and collision avoidance system (TCAS), as well. While TCAS explicitly coordinates evasive, tactical maneuvers to avoid traffic, the AOP offers maneuver options for both tactical and strategic traffic avoidance. Because it is designed to utilize broadcast data links, AOP relies on implicit coordination with other aircraft. Ballin hastens to add that the AOP is not intended to replace TCAS, which will keep aircraft apart in the rare case that a redundant AOP fails.

Wing explains how pilots might use the AOP strategically. "If a conflict alert pops up on the display, the pilot can turn to a special AOP page on the FMS control display unit and activate the "Resolve All" command. AOP then calculates an optimum [according to pilot's preprogrammed preferences] flight plan that avoids traffic, which automatically appears on the nav display in blue. When the pilot pushes a button to select the plan, it appears on the display in white, indicating that the FMS has received it from the AOP. The FMS would consider the white-line route, which factors in planned speeds and altitudes, to be a modified route.

"When the pilot pushes the `Execute' button, the flight plan appears in a magenta color," Wing adds. At this point, if the FMS is coupled to the autopilot, the aircraft follows the flight plan, and the transponder automatically broadcasts the aircraft's intent, referred to as the "command trajectory," to other aircraft and to controllers.

Should the pilot need to avoid weather, he can decouple the lateral and/or vertical navigation (LNAV/VNAV) modes of the autopilot, pushing buttons on the mode control panel below the glare shield. Clear of the weather, the pilot can then reactivate LNAV and/or VNAV, and the FMS will automatically direct the autopilot to make maneuvers back to the flight path.

For effective crew interface and to reduce the amount of crew training for AFM, NASA has made sure that the AOP imagery conforms to current flight displays in terms of color and symbology. For example, when the arc on the compass rose turns to an amber hue, it matches the normal color for pilot "caution" alerting.

The AOP algorithm also includes a "threat filter," which eliminates on the display aircraft that will not cause a conflict. The flight crew can turn the filter on or off and modify it to their convenience. For instance, they can filter out all aircraft that are nearby but flying away from their position.

The joint, Ames-Langley simulations represent the culmination of a seven-year-long series of evaluations designed to determine autonomous flight management's feasibility. NASA engineers believe they have proven AFM's feasibility. "AFM provides much growth opportunity, but much more needs to be done," says Ballin. "We need to fully understand the [ADS-B] messaging requirements and to conduct a safety analysis." A cost/benefit analysis also must be completed to determine the investment required by the airlines.

NASA also plans to have an airborne version of AOP installed in its Boeing 757 test bed aircraft. "We're looking initially to conducting [AFM] operations over the ocean, because the oceanic traffic flow problem is less complex [simply involving east/west tracks], and the operations may provide some immediate benefits for operators." says Ballin. "The oceanic region has less need for increased throughput, but it has no radar coverage, so separation is done procedurally, which leads to inefficiency for many flights. In oceanic operations we seek more fuel efficiency using AFM. The AOP will help pilots reach optimized positions on the oceanic tracks."

Receive the latest avionics news right to your inbox