WASHINGTON — The 20-month process of overhauling the flight software for the U.S. Air Force’s first dedicated Space Based Infrared System (SBIRS) missile warning satellite appears to have succeeded, but not before adding some $750 million to the overall cost of the program, according to an Air Force official. Meanwhile, ground testing on the satellite continues to uncover problems, the official said.
SBIRS prime contractor Lockheed Martin delivered the final version of the flight software for the geosynchronous-orbit satellite in June, and testing done so far has given no indication that further development will be needed, said Air Force Col. Roger Teague, Space Based Infrared Systems Wing commander at the Air Force Space and Missile Systems Center, Los Angeles. The software is slated to complete a test exit review by the end of July and shortly thereafter will begin qualification testing that should be completed next spring.
In a July 17 interview, Teague said the cost estimate for fixing the software problem is based solely on the amount of time it has delayed the program, which was well over budget and far behind schedule when the issue was first disclosed nearly two years ago. The cost of the SBIRS program, originally pegged at around $3 billion, is now in the neighborhood of $10 billion; launch of the first satellite, originally targeted for 2002, is now expected anywhere between late 2010 and spring 2011.
Lockheed Martin Space Systems of Sunnyvale, Calif., is under contract to build three geosynchronous-orbit SBIRS satellites and has received long-lead funding for a fourth; the Air Force has been directed to buy six satellites in total. Lockheed Martin already has delivered two SBIRS payloads that are hosted on classified satellites in highly elliptical orbits – both of those payloads are now operational – and is under contract for two more.
News of the SBIRS software problems surfaced via a September 2007 memo from then-Air Force Secretary Mike Wynne. The memo said the SBIRS satellites have a design similarity to an unidentified
U.S. government satellite that failed due to a faulty safe-hold mechanism, which refers to measures a satellite takes to protect itself from further damage when something goes wrong on orbit.
Teague said the SBIRS software problems were manifesting themselves in testing in any event. “In the previous version of the flight software, we were literally seeing hundreds and hundreds of [discrepancies] pile up at an alarming rate,” he said. “That’s when we knew that the design and architecture had fundamental flaws that needed to be addressed.”
While certain elements of the previous code were salvageable, the solution essentially required starting from scratch, Teague said. The final flight software that was delivered by Lockheed Martin will be used on all of the geosynchronous satellites.
“I’m very confident there will be no need for future software development changes, Teague said. “Based on the code quality we’re seeing and the performance in testing, we believe that software development activity was very successful and it will meet its mission requirements.”
While the SBIRS flight software was being overhauled, the Air Force took a calculated risk to continue with other testing and integration aspects of the first satellite while adding in blocks of software as they were completed. During these activities, a number of first-time hardware integration issues were encountered, Teague said.
“We’re seeing issues with a number of parts as we do our final integrated systems testing,” he said. “These issues are consistent with other new satellite platforms of similar complexity.”
Lockheed Martin spokesman Steve Tatum, in an e-mailed response to questions, declined to provide detail on specific component integration problems.
“Overall, we are progressing steadily through key integration and test activities on SBIRS, he said. “We are confident that we will successfully deliver this cutting-edge system while maintaining the mission assurance necessary for this critical national security program.”