Feature

By 2025, the friendly skies will be clogged with two to three times the present air traffic, conjuring up a host of problems and challenges, say federal experts.

To prepare, NASA, the Federal Aviation Administration and several related agencies are working on the Next Generation Air Transportation System, a congressionally mandated overhaul known as "NextGen." The plan promises to replace radar with satellite-based navigation and surveillance, a move that will vastly increase the amount and quality of advanced flight deck automation but still leave pilots in the driver's seat. And, in what could be a controversial change, NextGen proposes to give pilots more control of flight-path planning-thanks to the new automation and more sophisticated display designs-and removing much of that responsibility from air-traffic controllers.

With their expertise in human factors, NASA-funded psychologists and cognitive engineers are busy aiding this transition. In a variety of projects, they are exploring ways to prepare future pilots to cope with more information and make quicker decisions. They are also investigating the problems likely to arise among pilots, air-traffic controllers and the automated systems they'll use.

"The psychology of communications, of human-computer interactions with automated systems and of risky decision-making-all are important in this context," says NASA-funded aviation psychologist Christopher Wickens, PhD.

Stay out of my space

Psychologists at NASA Ames Research Center in Moffett Field, Calif., have been conducting a long-term program to address what will be a growing problem as air space becomes more crowded: maintaining a safe degree of separation between aircraft, called "separation assurance."

At a time when there will be three times the air traffic-and nine times the work managing it-the ability to maintain a safe separation among airplanes will be increasingly difficult, says experimental psychologist Walter Johnson, PhD, head of the Ames Flight Deck Display Research Lab. His team is studying automated systems that can calculate complete resolutions to such conflicts and then be digitally "uplinked" directly to the displays and automation on the flightdeck, substantially replacing human air-traffic controllers' verbal communications. The idea is to use air-traffic controllers on more of an "as-needed" basis, Johnson notes.

Although Johnson says it will take years before such systems are fully in place, an early version of an initial NextGen automated flight deck system is being tested by pilots of the shipping company UPS in collaboration with researchers from multiple organizations, including Vernol Battiste, a human factors expert at San Jose State University Research Foundation and a former air-traffic controller. UPS has already started equipping its fleet with these systems, and began flying aircraft with the equipment in January.

Given the differences in scale and operation of UPS and commercial aviation as a whole-UPS has a relatively small number of aircraft and flies mostly at night, for example- "the impact on the future of aviation of adding this equipment is a wait-and-see," says Battiste.

"But from the perspective of advances in flight-deck automation and flight-deck tools to support safe and efficient traffic management," he says, "it is a major leap forward."

Creating user-friendly displays

In related work, a team at NASA Langley in Hampton, Va., is examining ways to make cockpit displays more user-friendly, specifically examining how to reduce clutter and create more useful, visually sharp information, notes North Carolina State University cognitive engineer David Kaber, PhD.

Working with psychologists Amy Alexander, PhD, and Emily Stelzer, PhD, of Aptima Corp., the team is developing a model that can predict pilots' perceptions of clutter-both how they perceive the visual density of display features and to what extent they perceive their tasks to be hampered by irrelevant display information. Their findings will be used by aviation display manufacturers to improve display design, says Kaber.

The same team is also designing computer programs that will prevent errors when pilots "zone out" as a result of flying planes on autopilot for extended periods of time: Some planes include more than 15 kinds of autopilot control, and pilot attentiveness can vary depending on the type and length of autopilot being used, Kaber says.

Drawing from psychologists' research on working memory, long-term memory and decision-making, the team is developing a computer-based model of pilot behavior under various conditions, then checking it against pilots' actual behaviors.

One particularly intriguing form of future cockpit technology is called "adaptive automation," in which physiological sensors monitor a pilot's stress levels or cognitive state, and tailor the type and amount of automation and cockpit-display content accordingly. Depending on the pilot's state, as well as flight circumstances, automation may gather and present data on upcoming weather patterns, and pull together or suggest alternative routing for performance and safety.

For instance, if a pilot is entering a congested airspace and preparing to deal with other air traffic, physiological sensors may indicate high states of pilot arousal, leading to a "high work load" classification. In turn, the cockpit automation would present a simplified display of information that would give the pilot only the most salient information on avoiding a collision. Other more advanced forms of automation could suggest alternative flight routings for the pilot, displayed with similar clarity, Kaber says.

In a related project, Christopher Wickens of Alion Science and Technology in Boulder, Colo., and University of Illinois experimental psychologist Jason McCarley, PhD, are working on a NASA-funded project to address pilot "change blindness"-the common phenomenon in which people miss large, seemingly obvious changes in a given scene. Such work is critical given that today's automated alert systems don't always capture pilots' attention the way they should, Wickens explains. Some sound too frequently and too early, for example, creating a "cry-wolf" syndrome in which pilots learn to tune out even important signals.

One thing many NASA-funded projects have in common is that they use mathematical models to study the behavior of a complex system through computer simulation. These models are the wave of the future, Wickens says, because they allow designers to change parameters, enabling them to test many conditions without having to undertake elaborate real-world experiments. However, such programs have a long way to go before they accurately represent all aspects of a pilot's attention and awareness, says Kaber.

Much of this research is coming together in a prototype cockpit that the NASA-Ames flight deck team considers its crown jewel. Twelve years in the making and originally designed for free flight-that is, pilots designing their own routes as opposed to getting their information from controllers-the cockpit display group attempts to embed the "best" human factors principles, says Johnson, including the ability to display other aircraft in the vicinity and to detect and resolve potential conflicts. It also boasts a state-of-the-art cockpit display including graphical advances and flight automation to foster a high degree of "situation awareness" and reduce clutter, he says.

"We try and make its human factors features consistent with as many standards in the industry as we can," says Johnson. "But because it's so advanced, either we're addressing the problems the standards were trying to deal with in a different way, or applicable standards don't exist yet."


Tori DeAngelis is a writer in Syracuse, N.Y.