Engineers, airlines eye synthetic vision technological approaches
By J.R. Wilson
CEDAR RAPIDS, Iowa — Aerospace engineering could be called a "packrat" discipline, in which nothing is ever really thrown away, even when a program is canceled. Lessons learned, technologies advanced, research begun — all can come back in another form, another program, another time.
In the late 1980s and early `90s, leaders of Boeing, McDonnell Douglas, and NASA all were looking into the possibility of a replacement for the Concorde supersonic transport. The proposed new High Speed Civil Transport (HSCT) would be bigger, faster, and more comfortable than the old SST and would incorporate the latest advances in technology.
Today, everyone has pretty much lost interest in a new SST, but not in some of its research directions. HSCT engineers had been determined to eliminate the Concorde`s drooping nose, which the technological limitations of an earlier period had forced on their predecessors. The idea was to provide the cockpit crew with alternatives to out-the-window views.
From this effort rose two possibilities. The first is synthetic vision, which would be completely computer generated. The second is enhanced vision, which is sensor-based, and fuses everything from video inputs to infrared to millimeter wave radar.
"The high-speed research that was done at NASA to develop an SST without a drooping nose had a lot of elements of enhanced vision, with large-field-of-view displays instead of windows," notes Tim Etherington, a systems engineer at the Rockwell Collins Advanced Technology Center in Cedar Rapids, Iowa, and technical director on NASA`s synthetic vision information system.
By substituting large displays for windows "you can have sensor fusion and external cameras that give you vision without forward windows," Etherington says. "But in a normal cockpit, you`re better off just keeping the windows in an environment where the naked eye can work." NASA recently chose Rockwell Collins to lead a cost-shared research team to develop synthetic vision for airliners and business jets.
"NASA has indicated synthetic, rather than enhanced, vision would be a way to decrease the number of accidents on landing," Etherington says. "We will try to fold in pathway displays and 3D maps for the existing multifunction displays (MFDs).
To do this, a head-up display (HUD) with some type of terrain overlay is important, Etherington says. Engineers also are investigating new kinds of head-down displays. Experts from Flight Dynamics of Portland, Ore., are looking at the HUD, for example, and Collins engineers are strong in head-down displays."
Also on the Rockwell team are Jeppesen-Sanderson Inc. of Englewood, Colo., Boeing of Seattle, American Airlines of Dallas, Delft University of Technology in Delft, The Netherlands, and Embry-Riddle Aeronautical University in Daytona Beach, Fla.
"To realize a lot of the terrain capability and particularly any advanced 3D displays may require a new design," Etherington says. Liquid crystal displays (LCDs), he says, are promising as replacements to cathode ray tubes — particularly if engineers can design new LCDs with new graphic overlay capability. This approach, he explains, also would require a terrain database server of some kind on the aircraft.
Such a design probably will not pose an overly strenuous challenge for Collins engineers. "Most of our systems at Collins are being designed for dual use (commercial and military)," Etherington says. "Elements of the LCD displays, computing and processing resources and underlying architectures are designed to work with both markets. Our primary military market right now is transport, which is similar in many ways to commercial air transport needs." The Advanced Safety Program is directing the synthetic vision effort at NASA`s Langley Research Center in Hampton, Va.
"There absolutely is a safety and operational payoff for both military and commercial aviation," says Michael Lewis, Advanced Safety Program director. "In commercial, the biggest cause of delays is when airports go to instrument flight rules and capacity goes down. If aircraft in the future can operate with synthetic vision, we can upgrade the capacity of the entire system. On the military side, there is an issue of being able to get to where you want to go, especially with lesser-known locales. That also applies to some degree to tactical operations."
Lewis`s efforts aims primarily at so-called "controlled flight into terrain" — crashes that occur where confused pilots lose sight of where they are in relation to the Earth, and accidentally fly their aircraft into the ground. This phenomenon causes at least 30 percent of fatal aviation accidents worldwide.
"The premise I strongly support is that those types of accidents simply don`t occur in daytime visual conditions," he says. "So you take away one-third of worldwide fatal accidents if you can recreate those types of conditions, because humans are, by nature, geared toward out-the-window visualization."
The first step may be as an advisory display to address safety and the lessons learned from enhanced electronic ground proximity warning systems. By improving cockpit displays and adding operational experience, experts can open the way to applying synthetic vision techniques to approach and landing in poor visibility.
"The goal of our current research effort is to make a practical implementation possible within the next four or five years for either commercial or general aviation [GA] or both," Lewis says. "The development of an application of this type of technology is inevitable. This is how aviation will be flying in 25 years, not just GA or commercial or military, but everybody, because the benefits will be so great."
Government and industry experts are combining existing global positioning system (GPS) capabilities, which already tell pilots exactly where they are, with super-accurate terrain databases and graphical displays, including computer-drawn 3D moving scenes showing pilots exactly what is outside.
NASA leaders have committed $5.2 million, and industry leaders are committing $5.5 million for synthetic vision projects through the year 2000. Experts say more will come later to accelerate the commercial viability of synthetic vision, and make some systems available within four to six years.
One lingering question centers on exactly how to display the data. Military pilots say they have no problems with a helmet-mounted display — so long as it did not change the weight or shape of their existing flight helmets. Yet commercial, business, and general-aviation pilots do not wear helmets. Going to a full-glass cockpit design, as already is the case with many new aircraft, would provide one solution, but government and industry officials are looking for ways to retrofit existing aircraft with as few physical hardware changes as possible.
"If you look at the military application, with head-tracked weapons systems, you could head track the synthetic vision display as well. That would be the ultimate implementation of synthetic vision with interlooped aircraft control," Lewis says.
Advanced display technology is moving toward eyeglasses rather than helmets. Lewis says systems engineers can easily retrofit existing cockpits with glasses-based displays. Yet another alternative — wide-panel cockpit displays would be much more difficult, if not impossible, to retrofit.
"Some of the new GA aircraft are coming equipped with LCD systems, and there are manufacturers actively selling multifunction display units in the GA market. Those could be used, but to what operational capability could they be used reliably for an out-the-window synthetic display," Lewis asks.
Also crucial to making such a system work is the accuracy and extent of the terrain database, which experts would have to update constantly to incorporate everything from weather-related changes (such as snow) to construction work on runways to new buildings.
Clearly, the en-route database does not need to be as high resolution or as accurate as an approach and landing database, Lewis explains. As the aircraft gets closer and closer to the ground, however, the resolution requirements get higher and so does the importance of its accuracy.
"We envision nested databases, so near the airport the resolution and certitude of the terrain is there to higher and higher fidelity," he says. "It may be the control of the local database will be under the responsibility of the local airport authority. So given the requirements and standards for database development for whatever use is intended, each airport would be responsible to have the local map and environs updated as needed," Lewis says.
"Every daytime visual approach is an independent check of that database," he points out. "Even surprises would be seen hundreds of times during every daylight approach. Even beyond all that, perhaps an independent check is still needed and we would use an onboard system, such as a radar altimeter, to verify the database in real time."
Existing strobe cathode ray tubes on the majority of aircraft cannot properly display sufficient resolution or color for synthetic vision. One of the enabling technologies for synthetic vision is a raster LCD. The graphics horsepower required also may exceed the capabilities of standard avionics systems, requiring upgrades there to handle some of the envisioned displays.
But Rich Jauer, senior technical project engineer at the Boeing Phantom Works in St. Louis, says he believes no database by itself will ever suffice.
"Any system an aircraft flies will have to have some kind of external sensor," Jauer says. "You could have a perfect database today, but someone could dig a ditch or begin temporary repairs on a runway. So you need something to give the pilot a sense of what`s really out there (see story page 1)."
The enabling technologies available in the commercial market "far exceed any of the formats we envision doing, so we just need to catch up the avionics systems a bit," Etherington says. "We`re talking about a HUD with a wire frame representation of terrain."
Experts also are thinking about the role of "head-down" cockpit displays synthetic vision. "For the head down on the primary flight display, we`re talking about basically a `highway in the sky` type presentation, with flight path vector as a primary element of the display," Etherington says. "Even though the display will still have pitch and role and attitude behind it, the main elements of control will be flight path vector-based. It is more for awareness than to allow the pilot to use it to fly through terrain."
He says the concepts of the flight path and highway in the sky perspectives will enable pilots to operate aircraft in a more intuitive way than they do now. In addition, using procedures to enhance productivity in the terminal area could include more closely spaced runway approaches, eliminating some ground lighting requirements, and keeping the tempo of operations similar to visual flight rules even when visibilities decline.
"Airlines are now starting to fly into situations that are fairly terrain challenged. The performance limitations are being pushed right to the edge," Etherington says. "Airport approaches will get even more difficult as new noise restrictions are applied. A pathway display with terrain as a safety underlay will allow fairly complex approaches without an increase in pilot workload on the flight deck."
The first phase of the NASA contract will culminate at the end of next year with demonstrations aboard a NASA 757 of a HUD and head-down display incorporating weather, traffic and surface operations. In 2001 and 2002, follow-up awards will move the teams toward operational readiness rather than just concept demonstrations, Etherington says.
