By Dejan Damjanovic, Air and Marine Transportation
Solutions, Geoeye (www.geoeye.com),
Thornton, Colo.
Anyone who flies commercially likely is familiar
with flight delays. There are many complex reasons for such delays, but
the vast majority can be boiled down to a simple problem: too many
aircraft in too small of an area. Fortunately, new innovations in stereo
remote sensing are allowing aircraft flight controllers to minimize such
delays, thereby reducing flight times and fuel consumption as well as
increasing safety.
Why Do Delays Occur?
The aircraft congestion problem can be segregated into in-flight delays
that involve flying from airport to airport, and ground delays that
involve taxiing to and from the runway. To safely allow aircraft to move
on an airport surface or fly in and out of an airport’s vicinity, flight
controllers allocate buffer space to each plane, i.e., a containment
region, to avoid collisions. The faster an aircraft moves, the larger
the containment region must be. Moreover, aircraft cannot be in the same
3-D place at the same time as other aircraft; hence air traffic
congestion is a 4-D problem.
Aviation professionals speak of “airspace,” which represents a 3-D
polygonal area around an airport, starting at the ground or some
arbitrary altitude and extending up to a higher altitude. If one assumes
a 10-mile radius around an airport, then it follows that the buffer
areas of only so many aircraft can fit into that airspace per hour
before there’s no more room.
On the ground, it’s a much simpler version of the same problem. The
taxiways and runways of any given airport are only so long and so wide,
and each landing and return to the gate, as well as each departure from
the gate and taxi to the runway, takes a certain amount of time. Each
plane needs so much space reserved for it to do this safely.
Over the years, several ground-based radio devices have been
developed to help aircraft navigate safely. However, the limiting
problem was that each aircraft, ranging from a small Cessna to a
massive 747, used the same navigation tools. This approach results
in allocating the same containment region around each aircraft,
large or small. The ideal solution would be to define airspace
around airports in the most efficient way possible to allow the most
airplanes in and out, and to somehow reduce the necessary
containment region per aircraft so controllers could move more
airplanes per hour in and out of each airport. But reducing the
containment region reduces the separation between each aircraft,
which increases the risk of a collision. What to do?
A Better Way to Navigate
The advent of the Global Positioning System (GPS) constellation in
the late 1970s provided more precise 2-D navigation. With more
sophisticated altimeters known as “air-data computers,” aircraft
could better define their altitude to closer tolerances of ±
10 to 20 feet, instead of ± 200 to 300 feet, thereby increasing 3-D
navigation accuracy. Radar technology also improved, so air traffic
controllers could better locate where airplanes were in 4-D space
because airplanes could accurately report their own 3-D position.
Moreover, in the United States, the Federal Aviation Administration
(FAA) funded the development of the Wide Area Augmentation System (WAAS),
which used a series of ground-based stations and several
geosynchronous satellites to augment the precision of GPS
geolocation.
Such technology innovations led to the development of a new method
of air navigation known as required navigation performance (RNP),
which attempts to segregate airplanes by their ability to navigate
in 4-D airspace and allows airplanes that can support greater
accuracy to fly in smaller containment regions. If an aircraft can
fly in a smaller containment, it’s possible to get more planes into
and out of any given airport.
With RNP, each aircraft is assigned an index number that corresponds
to its ability to navigate accurately in 4-D. For example, as
detailed in the table above, if a pilot is flying a single-engine
airplane with a simple GPS receiver, he or she might be assigned an
RNP value of 2, or “RNP-2.” That means the pilot could determine the
aircraft’s position relative to Earth within two nautical miles, or
a containment region of four nautical miles. Contrast that with a
commercial airliner that has sophisticated flight management
computers, along with GPS and WAAS receivers. Such an aircraft would
be assigned “RNP-0.3,” because it can determine its position to
within 0.3 nautical miles. Therefore, an air traffic controller
could get six times the number of RNP-0.3 aircraft into the same
space as RNP-2 aircraft. Ultimately the idea is to define routes in
and out of an airport that segregate aircraft with good (small) RNP
numbers into high-volume, low-delay areas and other aircraft with
higher (poor) RNP numbers into low-volume, high-delay areas.
Another benefit is that air traffic controllers can define the
airspace into irregular polygons that are optimized for such
segregation. During the last 50 years, when airspace was defined by
early ground radio aids, airspace could only be defined in circular
shapes. The advent of GPS 2-D navigation and WAAS 3-D navigation
made it possible to define irregular polygons of airspace that
segregate good RNP aircraft from bad RNP aircraft. The FAA will be
redefining airspace over the United States during the next few
decades, as will other national aviation agencies in their own
countries.
Airport Surveying by Remote Sensing
To use optimized RNP routes in and out of airports requires 3-D
surveys of all obstacles that might be in the path of the new
routes. This requires stereo remote sensing to support 3-D feature
extraction of the following:
• Airport runways, taxiways, ramps, and buildings to accurately map
the airport and optimize taxiing routes (usually implemented in ESRI
Shapefiles).
• Obstacles that may be tall enough to pose a hazard to aircraft
taking off and landing on the new irregular RNP-based routes
(usually implemented in Shapefiles).
• Terrain and other natural features like tree lines that may be
tall enough to pose a hazard to aircraft taking off and landing on
the new irregular RNP-based routes (usually implemented in GeoTiff
or TIN formats).
The stereo source imagery can be from several sensors, including:
• single-orbit stereo satellites such as IKONOS, OrbView-3 and
QuickBird
• aerial stereo ortho-rectified image pairs coupled to GPS
• LiDAR or IfSAR image-collection systems
In areas such as the continental United States and most of the G-8
countries, all three of these imagery sources are readily available.
In many second- and third-world countries, or countries that
restrict GIS information, single-orbit stereo satellites become the
only option.
Once the imagery has been collected and
imported into a stereo photogrammetry tool such as SOCET SET from
BAE Systems (www.baesystems.com/gxp),
which seamlessly supports all three of the aforementioned formats,
it’s possible to collect the appropriate datasets. Then the three
building blocks—airport features, obstacle features and terrain
features—implemented as GIS databases can be used to create the
actual RNP procedures.
It’s essential that all three datasets are collected from the same
source and possess the same geolocation frame of reference. If one
used one set of source data for the airport and obstacle features
and another for the terrain, then the temporal/timing differences
would preclude getting a true, consistent 3-D view at a specific
point in time.
RNP Design and Benefits
To actually create an RNP route, one needs to have the three key
datasets previously mentioned: airport GIS model, obstacle model and
terrain model. Then the user must determine if he or she is
developing an arrival procedure—also known as a Standard Terminal
Arrival Route (STAR)—or a departure procedure—also known as Standard
Instrument Departure (SID).
A STAR wants to allow as many aircraft as
possible to arrive at the airport, and make sure that as many as
possible are allowed to depart the airport at the same time, via a
SID. This could be as simple as arriving from the north and
departing to the south, or it could be more complex. Next the user
begins to introduce more complex forms of geospatial data.
• Cadastral GIS data can be analyzed with housing densities in the
vicinity of the airport to see how to minimize noise generated from
flight operations.
• Habitat GIS data can be analyzed to avoid environmentally
sensitive areas.
• In a post 9/11 world, one may want to specifically avoid flights
over militarily sensitive areas or infrastructure such as nuclear
plants.
Through an RNP, users can combine all of the available GIS
information into more complex routes that benefit aviation users and
meet public requirements. A classic example of this is San Diego’s
recent selection process for a new airport location. Although
several sites were considered, the final selection used all of the
aforementioned GIS datasets to make a final choice. The table below
summarizes the findings and GIS data types used to reach the final
decision.
Thus, conventional GIS databases used to work with terrestrial
problems can be used to determine the higher flight volumes allowed
by RNP navigation routes. Of course, in most cases RNP’s biggest
benefit is fuel savings due to more efficient routes, which matters
greatly in a world where, at the time of this writing, crude oil was
selling for more than $70 per barrel. An RNP route is more efficient
because it allows an aircraft to “coast” down from higher altitudes
with little or no engine use, thus conserving fuel on every trip.
Solving the Next Problem
One of the other great changes in the aviation system of the future
will be the increased usage of unmanned aerial vehicles (UAVs)—robot
aircraft without pilots or crew that will fly aloft for 10, 15 or
even 20 hours at a time. The UAVs will be looking for terrorist
movements, illegal fishing fleets, environmental spills, illegal
immigrant movements and other serious challenges. The best way to
allow piloted aircraft and UAVs to operate in the same airspace is
to define special RNP procedures just for UAVs and to modify
traditional RNP procedures to avoid the UAV areas. In the
not-too-distant future there may be hundreds or even thousands of
UAVs in U.S. airspace. Stereo remote sensing and GIS databases will
help make a safer future possible
