U.S. Federal Aviation Administration (FAA) air traffic controllers at the Syracuse, N.Y., Terminal Radar Approach Control (TRACON) facility, along with their counterparts at the El Paso, Texas, TRACON, are now at the leading edge of the agency’s thrust to modernize the nation’s air traffic control (ATC) system. These two locations have been chosen as evaluation sites for the new Standard Terminal Automation Replacement System (STARS), a computer and display combination designed to not only handle the greatly increased traffic volume forecast in the decades ahead, but also to accommodate new technologies now in development for Free Flight.
Built by Raytheon Systems Co., Marlborough, Mass., under a $940-million contract awarded in 1996, STARS installations are planned for 191 FAA and 140 U.S. Air Force terminal control facilities throughout the National Airspace System (NAS). STARS has three major components: the control/display system and two computers, one of which is the primary data processing unit and the other the emergency back up system. The two computers are entirely independent of each other, and use totally different software programs to achieve the identical end function.
The primary unit’s development continues at Raytheon. The equipment installed at Syracuse and El Paso consists of the new high-resolution color displays that are 30% larger than the displays they replace, and the emergency backup computers, which are interfaced with the already existing, earlier generation, Automated Radar Terminal System (ARTS) computers. This combination, known as the early display configuration (EDC), is what controllers at the two TRACONs are evaluating.
STARS provides a highly reliable, open architecture design with significant built-in growth capability that allows for "easy and rapid incorporation of new data processing enhancements and new technologies planned for the terminal area," a Raytheon spokesperson tells Avionics Magazine.
When development of the primary computer system is complete–presently forecast for 2002–test units will be installed at Syracuse and El Paso to replace the ARTS computers, and the emergency computer will revert to its design role as a backup. At that point, the configuration will be known as Full STARS. However, it will still be subject to further evaluation until the controllers are satisfied that every in-service problem has been resolved. Then, and only then, can Full STARS installations commence throughout the NAS.
In Full Use
But the current EDC evaluations are not off-line test activities, nor will be those of the full system. At Syracuse and El Paso, the EDC systems are in continuous, 24 hours a day, seven days a week, operational use. Initially, the evaluations are aimed at getting comments from working controllers on the STARS display presentations, plus determine their ease of use and their impact on controller workload. Later, more advanced system features and components will be progressively added as the ongoing Raytheon development proceeds. This progression is in line with FAA’s new found mantra of "build a little, test a little." In 2001, 11 more EDC configured installations will be made at still unnamed FAA facilities, to broaden the evaluation base.
The display evaluation is of critical importance, being the key link between the controller and the terminal airspace outside. In the ATC environment, and particularly in the terminal area, the human factors of any man/machine interface become vital, with no room for errors or ambiguities. Consequently, only working controllers can assess whether the system’s design engineers have "got it right," and in the early years of the STARS development program, it was clear that the engineers hadn’t always achieved that ideal.
The STARS display is a major breakaway from previous technology used by the FAA. Since radar was first introduced in the U.S. for air traffic control, controllers have sat before large, circular radar displays, where a thin line of light, representing the rotating beam of the radar, sweeps steadily around the screen like the second hand of a clock– but faster, at one complete revolution every 4.7 seconds. When the radar beam strikes an airplane or other reflecting object, it causes a mark, or blip, to appear on the screen at the appropriate range and bearing.
Most ATC radars use two antennas: One transmits the primary radar pulses, which reflect back to create the blip or "skin paint." The other transmits the secondary surveillance radar (SSR) interrogation signals; it causes the aircraft’s transponder to transmit its flight data and altitude, which then appear beside the blip. On these older displays, the initial brightness of the blips and the SSR data slowly fades until refreshed by the next sweep.
STARS brings the radar into the 21st century, with a large, 20-by-20-inch (50.8-by-50.8-cm) color display screen where the blips and data remain at a constant, controller-selected, brightness level and where the radar beam’s sweep line is no longer visible. New display technology also provides a much sharper, cleaner picture, and along with it, enhanced capabilities for future applications. For the time being, however, the information on display is essentially identical to that provided on the earlier circular screens.
Got to Look Forward
So why the changeover? Simple, old technology. Parts of the current system’s software, for example, are written in computer languages that are, to today’s younger engineers, as incomprehensible as Latin. Equally important, the ARTS architecture is incapable of meeting future FAA requirements.
In its final configuration, STARS will accommodate such advanced Free Flight technologies as the Conflict Probe and the Final Approach Spacing Tool, both of which are now undergoing controller evaluation at other sites, but neither of which could be driven by the older ARTS computers.
No Casual Endorsements
As well, STARS breaks away from FAA’s traditional preference for custom built systems, and instead is designed around commercial-off-the-shelf (COTS) components, based on easily upgradeable open systems architecture. The color displays, for example, are standard, unmodified, business work stations, and the complete STARS system is, in fact, largely based on Raytheon’s similar AutoTrac and TracView equipment, which have been purchased by air traffic control authorities in many overseas nations.
How do working controllers like what they have seen so far? We posed that question during our visit to the Syracuse TRACON. We got two different answers: the official one and the unofficial one.
Officially–presumably because the FAA is still evaluating the system and is very sensitive about earlier controller dissatisfaction with STARS–local staff were not permitted to make comments for publication, unless they were first approved in Washington. But their unofficial answer was clear, in the very positive way the TRACON staff explained the system’s features and benefits, frequently referencing these to the earlier, circular ARTS displays.
Air traffic controllers are, perhaps wisely, a conservative group and not given to casual endorsements. The STARS situation was summed up by one of them–not from Syracuse or El Paso, but an old timer from another FAA facility. Having seen several previous "final solutions," he said, "The new display is certainly a real improvement. But we won’t know if the complete STARS is alright until we’ve had the full system running for a whole year, using all its features and under all traffic conditions, and every bug has been worked out of it. Then we can say it’s a good system. But not before."
Meantime, FAA and Raytheon have big money riding on their conviction that STARS will be the "good system" that the controllers have been awaiting.
A Bad Call on QWERTY
FAA’s original plan was to make the first STARS installation at Boston’s Logan airport in late 1998, just two years after the contract award. It was an ambitious schedule, intolerant of any delays. But minor software problems were starting to surface, and any activities not essential to meeting the target date were put aside.
One of these activities was the required human factors review–officially called the computer/human interface (CHI, pronounced Kye). Program officials felt they had a sufficient understanding of the controllers’ needs. As well, the basic Raytheon design was in wide and satisfactory use by controllers around the world. But in retrospect, bypassing CHI was a bad call.
In response, officials of the National Air Traffic Controllers Association (NATCA) promptly cited a number of what they felt were unacceptable features of STARS, and their battle of words with FAA commenced. Major shortcomings named by NATCA included the use of drop-down Windows-style data boxes, whose opaque backgrounds blotted out radar targets on the display screen behind them. Also cited: the standard, off-the-shelf computer keyboards in place of the special units used with the current ARTS computers. The first six letters of the new keyboards were in the conventional QWERTY sequence, but the first six letters of the ARTS keyboards were ABCDEF, and continued in alphabetical order to Z. NATCA stated these were critical CHI issues, but FAA was unmoved.
Only after Congressmen heard a NATCA representative state that the planned installation of STARS at Washington National airport would pose a safety threat did FAA feel forced to capitulate. Controllers became part of the STARS program team, which would, over a two-year period, list 98 separate CHI concerns, many of which are still being worked on, including the opaque windows. (The Syracuse EDC installation lacks windows, but does have the ABCDEF keyboard.)
Raytheon says it corrected the above problems long before systems at El Paso and Syracuse were installed, and "with the earnest cooperation of NATCA and PASS [Professional Airways Systems Specialists]."
For the STARS program, the FAA’s decisions formed a classic example of "more haste, less speed." The first installation at Boston was–and still is–postponed, as were several others, including Washington National. In fact, the entire program was restructured to accommodate the CHI requirements, with initial installations of partial STARS, i.e., the EDC package, being limited to the lower activity TRACONs in Syracuse and El Paso.
The 1998 date forecast for the first Full STARS installation has now slipped to 2002, and high profile Boston has given way to lesser known Syracuse and El Paso. As well, while the original STARS design called for some 120,000 lines of new computer code to adapt Raytheon’s international standard system to FAA requirements, the design changes now needed to meet the CHI demands have ballooned that number to more than 400,000 lines. And the final day of reckoning arrived in May, with FAA’s announcement of the amendment to Raytheon’s contract to accommodate CHI, which will add a cool $270 million to the STARS tab, now forecast to total $1.3 billion.
On the other hand, the U.S. Department of Defense (DoD) not being a NATCA client, decided to press on ahead with its half of the STARS procurement, and already has a Full STARS installation, sans CHI modifications, in operation at Eglin AFB, Fla. Raytheon says it, along with FAA and the DoD, recently completed operational testing of the baseline full service STARS system at Eglin. Raytheon adds that DoD intends to implement the baseline full service STARS, replacing aging equipment at its sites.
Like their FAA counterparts, of course, DoD’s working controllers don’t give personal interviews, but word from the top in the Pentagon is that, CHI or no CHI, they just love the system. When FAA and NATCA agree that all the CHI changes have been installed, evaluated, and are operationally acceptable, they will then be imported into the DoD equipment.
Meanwhile, back at the FAA, program managers are already memorizing a new rule: never, ever, sidestep CHI to meet a contract delivery date.
And Then There’s Ollie
No description of the STARS program would be complete without mentioning Ollie–the controllers’ preferred moniker for FAA’s formally named ARTS color display (ACD).
The ACD was conceived as a stopgap measure to stanch the bleeding of the NAS radar infrastructure, caused by the escalating failure rate of the early generation circular radar screens. At major air route traffic control centers (ARTCC), it is said, technicians were poised like emergency room interns, ready to quickly push a trolley loaded with a freshly serviced display unit to the site of the latest death.
Built by Lockheed Martin, the Ollie/ACD was designed as a modern display upgrade for the ARTS computers, pending the eventual total replacement of ARTS computers and displays by STARS, which was at that time still in the concept stage. But unlike STARS, the Ollie screen doesn’t–and was never intended to–do anything more than the older circular displays, except to be much more reliable.
Currently, FAA has ordered 295 ARTS displays, and installations will commence this summer, initially at control locations where failures of the older displays can least be tolerated.
Unfortunately for FAA, a perception seemed to grow among some controllers, who had seen ACD prototypes in operation, that it was the equivalent to the still distant STARS. Reportedly, this perception still exists in some quarters, and it continues to give senior FAA managers heartburn.
And why the name Ollie? Apparently, this was the code name for the initial development of the ACD concept, which was carried out somewhat covertly and without formal management approval. It’s along similar lines to the Iranian activities of U.S. Marine Col. Oliver North, who was then being keel-hauled by Congress for his undercover work. Uniquely, therefore, North’s name will go down in the histories of both the Marine Corps and the FAA.