Behind the STARS InterfaceBy David F. Carr | Posted 2002-04-09 Email Print
Online exclusive: Protests by the controllers union set development of the agency's Standard Terminal Automation Replacement System back by four years.
Behind the STARS Interface
At the FAA Technical Center in Atlantic City, the testing lab for STARS has rows of the new workstations set up side by side with several versions of ARTS. BASELINE found David Cognata, a retired controller who has been working on the STARS operational test and evaluation, sitting right next to a particularly archaic-looking ARTS console, like something out of a WWII movie. Although there are also newer ARTS consoles with color screens, this model's porthole-shaped screen, with faint amber dots lit by the sweep of the radar, made for a particularly sharp contrast with the bright color display of the STARS workstation.
The STARS workstation is a custom gray cabinet with a Sun Microsystems computer running Solaris 8 buried deep inside. The keyboard can actually be either ABC or QWERTY, as in the Department of Defense configuration. At NATCA's insistence, each workstation also has a row of knobs for specialized functions on either side of the monitor, mirroring those on the ARTS terminal.
Cognata demonstrates the utility of the knobs by showing how he can change the range of radar data displayed on his screen with a quick twist of the wrist. He can perform the same function by using the track ball to click a button on the on-screen button bar, entering the range in miles, and hitting enter, but it takes a few seconds longer. The knob gives him something more like the zoom control on a camera, making the concentric circles representing zones of radar coverage expand or contract with each click of the knob.
While it may be true that DOD was more willing to buy a system with a more PC-like user interface, he says, "the civilian controllers would never accept that because it requires too much data entry."
Bill Blackmer, the safety and technology director for NATCA, argues that the kid working behind the computerized cash register at Kentucky Fried Chicken doesn't work off a QWERTY keyboard, either. Instead, he has keys that say things like "Original Recipe" and "Extra Crispy." Besides, TRACON controllers don't sit at a keyboard and type. Instead, they occasionally hit a few keys required to accomplish a specific function, and they worried that changing the keyboard layout would slow them down.
It's not fair to write off the last couple of years' worth of development as being all about keyboards and knobs. In addition to pointing out how the color display makes it possible to pack more data onto the screen, Cognata showcases a number of flourishes added in the latest STARS releases that add valuable new functionality. For example, he can click on an aircraft and immediately call up a list of the closest airports where it could make an emergency landing. Or he can get a list of the hospitals it is closest tojust the thing for getting a medical helicopter safely to its destination.
Whether the user-interface changes should have been agreed to remains more a matter of management philosophy and politics than of technology. Agam N. Sinha, the vice president at Mitre CAASD who acted as a referee in the negotiation between the unions and the FAA, says a lot of the year-long process of reaching a compromise involved smoothing over the grudges each side held against the other. In addition to NATCA's complaints, the FAA also had to address another set from the Professional Airways Systems Specialists, the union for FAA employees who maintain not only its computers but other elements of the system, such as radar equipment.
"Were all of the issues that were on the table purely technical? No," Sinha says. "Were all the changes that were requested absolutely necessary? No."
On the other hand, there is some truth to the idea that throwing too many user interface changes at users who play such a high-pressure role is dangerous business, Sinha says. "A lot of it is second nature to them, and it's got to be second nature because of the complexity of the job." And no one wants to be in the position of forcing the controllers to accept a change that winds up causing an accident.
The success of STARS in other countries doesn't necessarily prove anything, Sinha says. For example, European air traffic control is more structured and predictable, he says. "They operate as if it was bad weather all the time." In other words, they don't make the same distinction the U.S. does between "visual flight rules" and the more conservative "instrument flight rules" that govern flights at night and through storms.
The good news is that once the FAA agreed to modify the system, it charted an efficient course toward that goal.
For STARS, the spiral development approach meant starting with the Early Display Configuration release (the user interface-only version), gathering controller feedback on that basis, and then offering some of those same controllers the opportunity to get some "back room" experience with a couple of interim releases of the full-service version before rolling out a complete replacement for ARTS.
This is a version of the spiral development methodology that the FAA has embraced, which is aimed at reducing risks and clarifying project goals through an iterative series of prototypes. "The spiral approach told us to do the back room last, and take on the biggest risks first," says Bill Voss, who runs the STARS program as director of the FAA's Terminal Business Service. "In this case, the biggest risk was the user interface."
Putting several versions into the field also helped to win over the controllers, Voss says. "User acceptability comes down to how much they trust you to be there with the next set of fixes. Otherwise, they're going to hold your release hostage until they get every change they might ever want to have," he says.