Why is it so expensive to shoot down a spy satellite?
By Michelle Tsai
A missile fired from a Navy warship on Wednesday night shot down a U.S. spy satellite that had been malfunctioning 130 miles above the Earth. The price tag for the endeavor has been pegged at upward of $30 million and even as high as $60 million, depending on the news report. Why did it cost so much money to shoot one missile?
They needed to reprogram the weapons. Once the orders were given on Jan. 4 to destroy the satellite, the Department of Defense had just a few weeks to outfit two Navy cruisersâ€”the USS Lake Erie and USS Decaturâ€”with rejiggered Aegis defense systems and a total of three SM-3 missiles. Only one missile was fired on Wednesday, but the other two had to be ready, in case a second or third attempt was needed. Since each SM-3 missile costs $9.5 million, the tab for the munitions alone adds up to almost $30 million.
Customizing the Aegis system and missiles for the satellite mission was a major expense. The technologies were originally designed to intercept ballistic missiles using heat sensors, but the spy satellite was cooler in temperature. To account for this difference, the three SM-3’s needed new software, hardware, and sensors, and the launching systems had to be given new sensors and software updates. The bulk of this task would have been assigned to high-priced contractorsâ€”like Raytheon, the maker of the missile, or Lockheed Martin, maker of the Aegis system. And it would have taken a large crew of engineers to rewrite the code, debug it, and test it over and over againâ€”all within three weeks. The stakes were also higher than they would be for a commercial software release, as the system had to work perfectly in a 10-second window; there was no opportunity to fix problems with software patches later on. (Before this week’s launch, the same anti-missile system had been successful on eight of 10 tries.)
Which begs the question, how did they manage this within the timeframe? What kind of software model did they follow? What was their internal structure to actually allow such things? Imagine this kind of productivity and quality within everyday software scenario! So many problems solved!