The Digital Sky
Tired of sitting on the runway? In a radical experiment that may provide a glimpse into the future of air traffic control, UPS is testing a satellite-based data communications system that could unclog the skies.
Every night between 11:30 p.m. and 2:30 a.m., nearly windowless jets with distinctive brown tails converge on Louisville, KY. One by one, about 90 aircraft in the United Parcel Service fleet land at the company’s distribution hub flanking the Louisville airport, disgorge some 600,000 parcels, reload and hit the sky again. The system is remarkably efficient, and it has helped to keep UPS aloft as the nation’s ninth largest airline.
But with the parcel delivery business expanding and the midnight skies over Louisville growing crowded, UPS is turning to new technology to compress arrivals and departures. In a radical experiment that may provide a glimpse into the future of air traffic control, UPS is embracing new satellite-based systems, hoping to wean itself off conventional radar-based technologies. Using the new digital tools, pilots would glance at cockpit displays showing their precise position, the positions of other UPS planes and a map of the airport and its runways-a display enabled by a combination of satellite positioning technology and digital datalinks between aircraft. Air traffic controllers would still run the show, but pilots would gain a tool to maintain more precise spacing on takeoff and landing.
If the UPS experiment works, says Dave Ford, a top Federal Aviation Administration official involved with the cargo airline’s initiative, it could provide a model for enhancing safety and efficiency in the nation’s overall air traffic control system. “One goal is to reduce runway incursions and accidents. We think this technology could help us in those areas. And we think there is a big link to efficiency,” he says.
Efficiency is definitely the spur for UPS. “We believe we can increase our throughput with the same airport infrastructure,” says company spokesman Ken Shapero. “If we can bring planes in faster or out faster, we can beat our competition.” UPS predicts the technology will yield a 20 percent capacity jump at Louisville. A reduction of 20 to 30 seconds between some landings and takeoffs could shave about a half hour from the company’s nightly sorting operation, a significant savings when your business hinges on delivering parcels on time. The numbers are so compelling that UPS is preparing to seek FAA approval later this year to use the system for approach and departure spacing at Louisville and is banding with other cargo carriers to push for even broader implementation.
The question now is whether what’s good for the cargo industry is also good for what aviation insiders jokingly call “self-loading cargo”-the traveling public. Can these satellite and datalink technologies help avert air-travel gridlock? In theory, they could keep airports functioning at full capacity in foggy weather, allow airplanes to land in pairs on closely spaced parallel runways, make possible more precise instrument landings and help airplanes avoid runway collisions. If this suite of technologies becomes widely available, “we in the [air traffic control] services industry may actually get ahead of the demand curve,” says Frank Marchilena, executive vice president of air traffic control giant Raytheon.
As any passenger at LaGuardia, O’Hare or Newark knows, the air traffic control system is lagging behind the demand curve now. Delays hit record levels in 1999 and 2000, and the problem promises to get worse. Last year, 670 million passengers flew in the United States; the FAA predicts 1 billion passengers will fly in 2010. According to FAA statistics, bad weather gets the lion’s share of blame for the air traffic control-related delays. But weather causes widespread havoc in part because today’s air traffic control systems are a patchwork of technologies built over the past half-century that are being stretched to the limit by the ever-increasing number of travelers. Radar technology has significantly improved since it was adapted for civilian air traffic control after World War II, but the basic procedure remains the same. Controllers herd airplanes along a limited number of radar-monitored “highways” in the sky. When weather is bad, controllers close some of the highways, creating traffic jams. Bad weather also prompts controllers to enforce larger buffer distances between airplanes, increasing delays.
During the 1990s, the advent of Global Positioning System (GPS) technology-in which precise locations can be fixed by triangulating signals from any of 24 military satellites-promised a new approach. Using GPS, pilots can determine their exact locations without relying on ground-based navigation beacons. Over the past decade, a collaboration of several government-funded labs, including Bedford, MA-based MITRE and MIT’s Lincoln Laboratory, developed a new way to continuously transmit digital GPS position information and other digital data among airplanes and controllers. With this network of digital information (known to insiders as ADS-B, or Automatic Dependent Surveillance-Broadcast), planes can continuously exchange data on location, speed, flight plan, aircraft size and type, number of passengers and weather.
The system can be thought of as the latest-generation phone lines and modems in an emerging “Aviation Internet”-one term being used to describe the increasing data flow among planes, controllers, ground crews and aircraft maintenance facilities. UPS Aviation Technologies is the only company that has developed cockpit displays certified by the FAA to receive and display information from this newly developed datalink, but others-including GE-Honeywell, Rockwell Collins and L-3 Communications-are working on their own systems.
A fundamental shift to satellite-based tools would require a monumental effort to achieve consensus among pilots, controllers and regulators. How could consensus be achieved? In air traffic control, “things tend to be reactive rather than proactive, and that’s probably what’s going to happen here,” says Jim Kuchar, associate professor of aeronautics and astronautics at MIT. “A systemwide change is either going to occur because of a major congestion problem, or because efforts like UPS make it more attractive. If UPS gets this thing working and it shows all these benefits, maybe others will say, ‘we’ll take another look at this.’”
That second look, however, may be slow in coming. Four years ago, United Airlines pilot Rocky Stone proposed using the new satellite-based technology to fight congestion by allowing paired landings during poor visibility at the notoriously fog-bound San Francisco airport, where the runways are a whisker-close 250 meters apart. But the idea proved impractical in the short term, says Dave Jones, who directs United’s efforts to improve efficiency at its San Francisco hub. To implement the strategy, United realized it would need Boeing and Airbus to approve new cockpit displays, pilots and controllers to accept them, and the FAA to certify equipment and applications. And even if United had installed the system, its airplanes would still have had to get in line with other airplanes lacking the technology. In the face of these obstacles, the airline shelved the plan and is exploring advanced radar-based tools and procedures instead.
United’s experience illustrated a fundamental difficulty in implementing this new technology: it’s an “all or nothing” proposition. Unless all airplanes around a given airport are equipped with it, the system can’t be relied upon for spacing, collision avoidance or much else. “There’s got to be a whole architecture of the airspace that everybody has got to agree to,” says Robert Rosen of NASA’s Ames Research Center in Moffett Field, CA. “None of that is in place today. It [ADS-B] is kind of like a piece of the puzzle, and it may even be a cornerstone of it. But having it in place is still far from having solved many of our problems.”
The case for satellite tools is far more compelling where radar infrastructure is spotty or nonexistent-and where the safety benefits are obvious. One such place is Alaska’s 260,000 square kilometer Yukon-Kuskokwim Delta region, where small-airplane deliveries and transportation are an essential way of life-and death. Much of Alaska has no radar coverage, no air traffic control towers and no paved runways (gravel airstrips are a luxury), making the area more like remote regions of Africa or China than the lower 48 states. The 1990s saw an average of one aircraft accident in Alaska every other day, including 186 fatal crashes leaving 398 people dead. During that decade, Alaska accounted for 37 percent of the nation’s total aircraft accidents and 20 percent of total air-crash deaths.
In 1998 this carnage prompted Congress to appropriate $11 million to install new equipment in 155 small airplanes in Alaska. UPS Aviation Technologies provided the avionics, and now the GPS-based system is being used by Anchorage-based air traffic controllers to guide small aircraft in the remote, marshy delta. And while radar-devoid countries like Australia and even Mongolia are starting to deploy satellite air-traffic tools, the Alaskan region is the first place in the United States-and the only one in the foreseeable future-to move to 100 percent satellite-based air traffic control. (In the UPS experiment in Louisville, radar would still guide aircraft to and from airports. The satellite tools would only aid approach and departure spacing.)
Whether the technology can make similar inroads elsewhere is less clear. One concern has been whether GPS satellite signals are robust and reliable enough to serve as a foundation for air traffic control. But a 1999 report by the Johns Hopkins Applied Physics Laboratory helped allay fears that satellite signals-weak compared to ground-based radar-are at risk of disruptions from solar radiation, atmospheric disturbances or terrorist hackers. “Technologies are emerging that can greatly reduce vulnerability to GPS signal jamming,” according to the study. And to the extent that satellite signals are warped by the atmosphere or other interference, they can be verified and tweaked for extra accuracy with ground-based augmentation systems like those being installed by Raytheon.
Another fundamental question is whether new displays in the cockpit might distract pilots, and whether new navigation responsibilities will overload them. “Human error is involved in at least 80 percent of all accidents and incidents in aviation,” says Kim Cardosi, manager of “human factors” programs at the U.S. Department of Transportation’s Volpe Center in Cambridge, MA. “The work environment is so complex, it can set them up to make mistakes, and that’s what we have to guard against in these systems and displays. We have to make sure [pilots] aren’t overwhelmed by information, and that when they do make a mistake it can be corrected before it has serious consequences.”
The infusion of data and displays brings with it new sources of confusion. Kuchar cites “a number of accidents in which there has been a mismatch between what the computer was thinking and what the human was thinking,” such as the 1995 crash of an American Airlines Boeing 757 on approach to Cali, Colombia. The airplane slammed into a mountain, killing 160 people, when the autopilot was instructed to fly toward a radar beacon the pilot thought was near Cali, but which was actually near Bogot. In this case, radar beacon technology and the autopilot system helped lead passengers to their deaths after a seemingly trivial pilot error. “If the United States switches to widespread use of [satellite-based technology], there will be other areas in the world that haven’t, requiring pilots to use different procedures in different places. This can lead to additional errors and problems,” Kuchar warns.
Despite these questions, prototypes are advancing. A key day for UPS’s effort came last October, when FAA administrator Jane Garvey flew into Louisville for a joint FAA/UPS evaluation of the technology. Garvey stepped inside a UPS Boeing 727 whose fuselage had none of the usual furnishings-just several pods of computer equipment in the front, and 16 leather-upholstered first-class seats bolted to the floor in the rear. Taking a seat in the front row, she glanced at a computer monitor displaying a graphical depiction of the Louisville airport. The runways were dotted with slow-moving brown triangles. These represented airplanes equipped with the datalink system, exchanging position data with each other. “This is cool,” Garvey said.
“Wait until you see the in-flight movie,” quipped George Cooley, an engineer at UPS Aviation Technologies. The “movie” began as the 727 rolled slowly along the airport’s taxiways. Other taxiing airplanes were plainly visible on the screen. Suddenly, a blue triangle appeared on the screen, its tip elongated with a needle-nose indicating high speed. A moment later, the blue blip turned brown. Jim McDaniel, head of the FAA’s technology assessment programs, announced that an airplane had just taken off. But in fact, blue denoted an airborne airplane, brown an airplane on the tarmac. A moment later, he corrected himself. “I thought it was taking off at that moment, but it was landing,” he said.
The demonstration was only intended to show how the technology could augment runway awareness, and its basic benefits were apparent. Even in foggy weather, the cockpit display would have given a clear view of runway traffic and immediately made apparent any wrong turn. The system worked perfectly. On the other hand, its interpreter-in this case, an experienced FAA official-had briefly been confused by the display. The mistake was an apt demonstration of why approval of new technologies for air traffic control takes time: to ensure all sources of confusion have been ferreted out.
As UPS pushes its case, a consensus is emerging that growing demand will force changes in the nation’s air traffic control system. In the short run, some relief may come from airport expansion and new construction; some airports are also considering higher peak-period landing fees to discourage the rush-hour crush. Late last year, the FAA announced a lottery system for the assignment of flight times at LaGuardia-which by itself accounts for about one-quarter of the nation’s delays-to reduce congestion. New procedures and uses of radar tools are increasing capacity at airports like Dallas-Fort Worth. The FAA, for its part, notes that even in busy cities the system has plenty of capacity at off-peak times. “All of the technologies that we are working on address a piece of the pie, and together they will ultimately create more capacity, but it is going to be incremental, at best,” says Kathryn Creedy, an FAA spokeswoman.
However, NASA’s Rosen projects that the FAA’s incremental approach will only keep pace with demand for the next decade or so. “Because of the demand on the system, all the technology developers are focusing on next-generation tools,” Rosen says. “But we recognize that even after all these tools are in place and working together, demand is such that it would soon again exceed capacity.”
The basic technologies for satellite-based air traffic control-the GPS system, datalinks, computational power and compact cockpit displays-are on hand. But there is nothing close to consensus on how-and whether-to deploy them widely. So far, the public outcry hasn’t been loud enough, the airlines haven’t seen the business case, and the FAA hasn’t tried to force a systemwide change. “Gridlock is in the eye of the beholder,” Rosen says. “However, everyone agrees that it will get worse before it gets better.”
Last October, UPS got good news: the company received the first FAA certification for its new cockpit device. It was only a small step and only for a very limited purpose, to help pilots gain “enhanced see and avoid” capability in the skies over Louisville. But the approval did signal that the system is making its way onto the regulatory radar screen. “There are still a lot of challenges with this-a lot to be resolved,” McDaniel says. “The pilots and air traffic controllers are excited; there’s a lot of potential, but they are not the least bit bashful about what it needs-for instance, [reducing] the clutter on the screen.”
If those problems can be ironed out, however, the UPS initiative at Louisville just may be the first step in helping reduce clutter in the skies.
Become an MIT Technology Review Insider for in-depth analysis and unparalleled perspective.Subscribe today