DOT bashes Stars, calls on FAA to stem runaway cost

 - October 8, 2007, 10:39 AM

In a September 9 report to the FAA Administrator, the DOT’s inspector general called upon the agency “to reevaluate the costs of Stars [the standard terminal automation and replacement system] and consider other alternatives.”

Stars was envisioned by FAA planners as the centerpiece of the agency’s terminal automation modernization strategy, which would replace aging 1970s-era equipment at every ATC terminal in the NAS. But as the IG’s report clearly illustrates, this promise has not been, and may never be, realized.

The 1996 development, production and installation contract to Raytheon covered 188 Stars controller consoles and was to cost $940 million, with all equipment delivered by 2005. Three years later, in 1999, FAA program managers had revised their 1996 figures, and predicted that the cost would rise to $1.4 billion, with deliveries extended to 2008. And another three years later, in March last year, the program managers revised their figures upwards again, forecasting that the cost would be $1.69 billion (almost double the original contract value), and delivery completion would be seven years late, in 2012.

Yet in 2001, FAA officials testified to Congress that Stars development was complete. The IG’s report noted that since then the agency has spent more than $170 million on further development, including “a significant amount of work” correcting problems encountered at the flagship installation at the Philadelphia Tracon. To date, only a few Stars installations have been delivered and none, including PHL, meet the requirements of major centers, according to the IG. Z

FAA program managers were also required to forecast Stars’ life-cycle costs, including development, production, installation, hardware/ software updates and operations. Life-cycle cost estimates are essential to the procurement process, since they show how new acquisitions will affect an organization’s purchasing budgets over future years. In 1996 Stars life-cycle costs were forecast to be $2.9 billion. In 1999 this figure was revised upward to $4.2 billion. By 2002 it had reached $6.1 billion–more than doubling the original estimate.

Program officials were unable to provide the IG’s staff with any forecast data later than March 2002. But they did claim that they planned to save $281 million, partly by incorporating $35 million in technical improvements (but which cannot be verified as acceptably safe until this month) and partly by “saving” $147 million by simply charging it to another FAA account.

The report also criticized the way in which program officials were monitoring program costs, or, in the case of Stars, not monitoring them. Over the program’s seven-year history, the contractor’s invoices–currently totaling $688 million–have never been independently audited, and FAA officials were unable to explain a $41 million difference between what Raytheon had billed the FAA and what their records showed as work actually performed.

How, then, has ATC functioned without Stars? Because of Raytheon’s inability to deliver, the FAA was forced to purchase off-the-shelf equipment from Lockheed Martin, and these units are now operating in 141 terminal facilities across the continental U.S., including 10 of the largest. Called Common ARTS (for automated radar terminal system) the equipment was shown by the FAA’s own analysis to have more than 90 functions that Stars could not offer, and which would require further development, for unknown additional cost, to incorporate.

In their defense, Stars program officials maintained their system had unique capabilities that did not exist in Common ARTS, but the IG’s investigators did not see evidence of this. They were, however, assured by program officials that the manual control knobs on the Stars display made it more desirable than the Common ARTS display, which lacks them. Separately, however, the FAA reported having successfully fielded “hundreds” of these knobless units the Atlanta, Dallas, New York, Northern California and Southern California Tracons and, late last year, at its new consolidated Potomac Tracon.

Yet the most remarkable revelation in the IG’s report is that the Stars program officials propose that when their system reaches full operational readiness–at an unspecified date, and unspecified cost–it should replace every Common ARTS installation currently in use throughout the NAS. Last month program officials planned to present their proposal, along with updated cost estimates, to the FAA’s top-level Joint Resource Council.

In the IG report’s words, the Stars program “is not the same program that was planned eight years ago.” The report, available at, appears to be a textbook example of the FAA’s historic inability to manage major programs, and underlines Administrator Blakey’s drive to instill business-like, performance-based practices in the agency, along with individual accountability.