Reprinted from the A&M
System Engineering website
by Lesley Kriewald
Science Writer
Texas Engineering Experiment Station
A
student tests the Texas A&M team's software in a flight simulator.
Wallace
Kelly (left) and John Valasek with the RTI Piper Aztec research
airplane at the 2005 SATS Flight Demonstration in Danville, Va., in
June.
(College Station)—An initiative aimed at getting more people
to travel via private plane for longer trips is gaining momentum, thanks
to the U.S. Congress, NASA and, in part, Texas A&M aerospace engineering
associate professor Dr. John Valasek.
Valasek is director of the Department of Aerospace Engineering's Flight Simulation
Laboratory and led a team of Texas A&M Engineering aerospace engineering
and computer science researchers to develop and evaluate algorithms, software
and displays that use artificial intelligence to help pilots of small, private
aircraft safely use airports without air-traffic control towers and radar
(called non-controlled airports) in near all-weather conditions.
The research was part of the Small Aircraft Transportation System (SATS), a
five-year research program funded by Congress and administered by NASA Langley
Research Center. Six SATS labs around the country each contributed to the project,
which seeks to enable a safe travel alternative to driving; open up access
to the many airports in small communities, thereby reducing congestion at major
airports; and invigorate the general aviation industry.
Valasek said that only 22 percent of the population of the United States lives
within 30 minutes of a major hub airport, such as Chicago O'Hare International
Airport or the Dallas/Fort Worth International Airport, while 41 percent live
within 30 minutes of a regional airport such as Bryan-College Station's Easterwood
Airport. But 93 percent of the people in the United States live near a SATS-type
airport, a figure that demonstrates the potential these airports have for general
aviation and air travel.
"Our motivation is to get people thinking about general aviation as an alternative
to other types of transportation for longer trips, including commercial flights," Valasek
said. "We want to make this affordable, convenient, and safe enough to
get you away from your car for medium-to-long trips."
To that end, the Texas A&M team developed algorithms, software and displays
to help those piloting small aircraft maintain situational awareness in the
approach and landing pattern, the most challenging portions of a flight.
The group applied artificial intelligence (AI) techniques to cockpit displays
and pilot aids.
AI is used here to assist with decision making in the cockpit and to anticipate
problems before they occur, Valasek said, and the flight segment identifier
is the realization of this approach. The flight segment identifier looks at
pilot actions and current flight conditions to infer what the pilot is attempting
to do and decides whether the pilot is conforming to procedures and operating
within proper airspace corridors. It does not control the airplane, although
it could. The pilot always has full authority in this system.
"Let's say you're flying along, and just like a car slowly drifting from
side to side in a lane, your altitude gradually drops," Valasek said. "Is
this because you're not holding your assigned altitude very well, or because
you are starting a descent? In a different situation, if you're flying very
fast and close to the ground with your landing gear up, are you attempting
a landing and simply forgot to lower the landing gear, or are you in fact just
flying close to the ground? The flight segment identifier evaluates the available
information and asks, `In my judgment you are performing this particular action.
Is this what you intended to do?' The pilot can then accept this judgment and
use it to make a decision and take corrective action, or disregard it and continue
with the original action."
The system's virtual instructor pilot makes sure that all procedures are
being adhered to and indicates whether or not a pilot is following a reasonable
sequence of flight segments. It displays both procedural information -- "this
is the segment you should be in" -- and state information -- "this
is the segment you are operating in" -- with concise text messages to
the pilot. Exactly what is displayed depends on whether the procedural and
state-based results match or at least follow a reasonable progression.
"Under normal operating conditions messages are displayed to the pilot in
green," Valasek said. "But if a mismatch occurs between what you
should be doing and what you are actually doing, the message changes color
to yellow to alert the pilot to take appropriate action."
A bonus feature of this approach is the ability to determine who is flying
an airplane. Over time, the system learns the characteristics of particular
pilots that fly that aircraft. Potentially, if an unauthorized person -- or
a terrorist -- takes over the stick, the system knows that an authorized pilot
is not flying and can take appropriate measures, such as go into a holding
pattern or alert authorities.
Valasek's Texas A&M team consisted of 13 undergraduate and graduate aerospace
engineering and computer science students. Computer science graduate student
Steve Wollkind and undergraduate students Arnold Binas and Carol Tsai were
essential in developing a synthetic traffic display called the Multi-Agent
Intelligent Distributed Airspace Simulation (MIDAS) so that pilots evaluating
the technologies would experience a simulated real-life situation of the
air traffic environment.
Aerospace engineering Ph.D. candidate Jie Rong was the lead graduate student
on the project, coordinating the work between the Texas A&M team and the
RTI International team for the cockpit display called RTI View. His dissertation
as well as that of Ph.D. candidate Martin Ding's was based on the SATS project.
Undergraduate aerospace engineering major and certified flight instructor Kyle
Helbing used the project as his Engineering Scholars Program research, and
then entered his work in the American Institute of Aeronautics and Astronautics
(AIAA) Region IV Student Paper Conference, where he won second place. Helbing
assisted Ph.D. candidate James Doebbler in creating a SATS capable multi-pilot
simulation facility in Texas A&M's Bright Building, and then designed
the approach plates and scenarios used to evaluate the system with Ph.D.
candidate Theresa Spaeth, who performed the human factors and pilot workload
evaluations. Besides Helbing, graduate students Tom Wagner, Paul Gesting
and Celine Kluzer, and undergraduate students Zach Reeder and Carolina Restrepo
served as evaluation pilots.
"Here at Texas A&M, we conducted one of the very first piloted simulation
evaluations of the SATS self-controlled airport concept," Valasek said. "We
had multiple pilots `flying' simultaneously in congested airspace around a
self-controlled airport, focusing on collision avoidance using advanced cockpit
displays, and real-time human factors such as pilot workload and situational
awareness. It was a complete success and nicely demonstrated the SATS High
Volume Operations concept."
Helbing added, "With the advanced aviation technology now available,
an almost endless supply of information is available in the cockpit. The
trick is supplying the pilot with just the right amount of information at
the instant they need it, and our talented group of research pilots helped
us in making those determinations."
The Texas A&M researchers were part of the North Carolina and Upper Great
Plans SATS team lead by RTI International and focused on the "High Volume
Operations" aspect of the program. Working closely with the Texas A&M
team was a group from Blue Rock R&D, a company formed by Texas A&M
electrical engineering graduate Dr. Wallace Kelly. Kelly earned his Ph.D. in
electrical engineering in 1997, and his dissertation, "Hypertrapezoidal
Fuzzy Membership Functions for Decision Aiding," formed the starting point
for both the Texas A&M and Blue Rock teams' work. Kelly is the co-holder
of a patent for Hypertrapezoidal Fuzzy Membership Function along with his Ph.D.
committee chair, retired electrical engineering and aerospace engineering professor
Dr. John Painter. Kelly also hired Painter as a consultant to assist on the
project with Blue Rock R&D.
The culmination of the five-year program was a public technology demonstration
by all six SATS Labs for NASA and Congress in June in Danville, Va.
"Perhaps the best aspect of this program for us was the flight demonstration," Valasek
said. "The vast majority of aviation research never makes it to the flight
stage. To see our work selected and then implemented for flight after all those
months of development and testing was very rewarding. It gave us the theory-computation-experiment
closure that all researchers seek."
Valasek said that the general public will not be the only ones to benefit from
SATS.
"Studies conducted by NASA using the SATS approach have shown that business
travelers stand to benefit a great deal too, particularly those who travel
regularly to medium- and smaller-sized cities not served by hub or spoke airports.
In fact, at least one company is looking into being a kind of SATS air taxi service
for the business traveler."