RENE MONTAGNE, HOST:
It's easy to understand how thrilled scientists and engineers felt a year ago when NASA's robotic rover landed on Mars. One missed calculation, one broken bolt, one wrong line of computer code, and after traveling 350 million miles, your precision spacecraft can either miss Mars entirely or land so hard it becomes a pile of useless junk. So when the Curiosity rover radioed mission control that it had landed safely on Mars, here's what happened.
(SOUNDBITE OF CROWD CHEERING)
MONTAGNE: True, the landing was perfect, but what the public did not know was that a last minute problem cropped up that had mission engineers biting their nails. NPR's Joe Palca has been exploring how scientists and inventors do their work in what we're calling Joe's Big Idea. Here Joe reveals for the first time details of the 11th hour dilemma and how NASA engineers dealt with it.
JOE PALCA, BYLINE: There's a box inside the rover Curiosity that tells the onboard navigation system how fast the rover's going and what direction it's headed in. These measurements are so precise that the navigation software needs to know exactly where the box is inside the spacecraft. That location effects all the other calculations the navigation software makes.
Now, just four days before landing, mission manager Peter Theisinger says engineers at NASA's jet propulsion laboratory in Pasadena discovered something disturbing. The navigation system had slightly wrong information about the box's location inside Curiosity.
PETER THEISINGER: We discovered that that parameter was not correct, that it was a few inches off.
PALCA: A few inches might not sound like much, but no one really knew how it would affect the rover's ability to navigate to the right place on Mars.
THEISINGER: People started to do a set of analysis about whether that affected EDL in any significant way.
PALCA: EDL stands for entry, descent and landing. Miguel San Martin was one of the senior engineers on the landing team. He says the straightforward option was to send the navigation software the correct parameter.
MIGUEL SAN MARTIN: We had the tools to change it easily. The problem is any touch on the spacecraft is a potential problem.
PALCA: Engineers hate, hate, hate to make last minute changes to software. Once you've thoroughly tested something in one configuration, you don't want to make any changes. You never know what unintended problems you might cause. So San Martin and his colleagues ran all the tests they could think of to try to figure out what would happen if they left the wrong parameter in the navigation software.
MARTIN: All the analysis indicated that this was not a problem. However, everything else that we've done we've been thinking about for years, and now we have a few days to think about this particular problem.
PALCA: Before any Mars landing, anxiety grips mission control. This problem just made things worse.
MARTIN: A lot of lost sleep during that time. A lot of tension, so that was probably the hardest time of the project for me.
PALCA: Ultimately the decision whether to send the correct information to the rover fell to Pete Theisinger. On Saturday morning, the day before landing, he scheduled a meeting.
THEISINGER: I think we started around 7:00 Pasadena time to basically decide what we were going to do.
PALCA: About 40 people crowded into a conference room at the mission operation center. Theisinger went around the room, asking landing team members whether they felt the command with the right information about the box should be sent.
THEISINGER: Some people felt we ought to do this clearly and some people felt that maybe yes, maybe no. Some people felt we shouldn't do it, maybe yes, maybe no, and some people felt we shouldn't do it clearly. And so it kind of came down to me and I decided that we should send the command.
PALCA: With no certainty from his team, he decided to risk changing what had been planned and meticulously tested.
THEISINGER: I think it was probably one of the tougher decisions I had to make because it was so even.
PALCA: The meeting ended by 8:30. By noon Curiosity had the new command. Now all engineers could do was wait.
THEISINGER: And I was thinking as went home that day that I was extremely confident that EDL the following night was going to work just fine.
PALCA: And as everybody now knows, it did. All the parts the landing team spent nearly a decade testing and tweaking, testing and tweaking worked just as they were designed to. Miguel San Martin says after the landing, the rover sent back data it recorded as it descended to Mars, data the engineers gobbled up.
MARTIN: It is a lot of fun to go through that after years of wondering, to go and say, oh look, this works just as expected. This was worse. This was better. In most cases it was better.
PALCA: For an engineer, that's the icing on the cake. Joe Palca, NPR News.
NPR transcripts are created on a rush deadline by a contractor for NPR, and accuracy and availability may vary. This text may not be in its final form and may be updated or revised in the future. Please be aware that the authoritative record of NPR's programming is the audio.