2009-03-28

Spirit Sol 83

We might start driving tomorrow, sooner than I thought -- if not tomorrow, then the sol after. They still haven't decided where we're going from here, along the rim or straight away from the crater. I start looking through the imagery we have, to see if we've got a safe path in either direction. Neither one looks good.

Opportunity is driving today. Only a few sols out of their crater and they're going to break our one-sol drive-distance record: 48m tomorrow, or so. And they'll better that record on future sols. Before long, they'll set a record we have no hope of breaking. Oh, well. It was nice while it lasted.

I chat with Frank about this for a little while, and about other things. They're going to stay on Mars time a little longer than we are. Frank isn't going to miss Mars time, but I am. I realized one of the reasons why when I was walking in today at 3:30 AM: coming to work at all times of the day and night emphasizes the uniqueness of what we're doing. When we start keeping regular, Earth-time hours, it just won't seem quite as special any more. Frank grudgingly agrees.

"You want to see something really cool?" Justin Maki asks. Who wouldn't? Justin shows us two really cool things. First is an animation of all the MI images that have been taken on Opportunity. The second is even cooler: Stubbe Hviid has put together a movie of all of the front HAZCAM images we've taken on Spirit. The movie incorporates all of our drives and IDD work up to this point -- the last three months of my life, flashing before my eyes. We've packed a lot into those three months.

Mark has solved the visual odometry problem that caused Opportunity to behave so strangely when they tried it. It turns out that the code wasn't properly thinking in 3-D, causing it to think the goal point was closer than it really is. We might get a solution in place and tested in time for us to use it when driving away from here, though I doubt it.

Today's sequencing should be simple: a RAT placement, a bit of IDD work. Of course, it doesn't work out that way. The latest HAZCAM mesh doesn't match up with the old one we've been using; it shows the rock face about 2cm farther away, so that our targets are floating in the air. A third mesh -- this one from yestersol's NAVCAMs, taken on the same sol as the new HAZCAMs -- agrees with the first one. So we end up using the old one, but not until we spend more than two hours arguing about it, comparing older images with newer ones to see if the rover really has shifted position (say, because we slipped off a rock while we were RATting), etc. And in the end we're still not a hundred percent sure whether we made the right choice. If we didn't, tomorrow's sequence will fault out and we'll be here another sol at least. We also don't know why the meshes would have differed, which is disturbing. Those images are critical to our ability to get work done; if we can't trust them, we can't safely operate the rovers.

I get the sequence more or less done and hand it over to Bob, then spend the rest of my time (plus a couple of hours more) planning our exit from here. Meanwhile, "Angry Bob" earns his nickname[1], swearing louder and louder at the software as he finishes the sequencing work. I can't help laughing, but I try not to let him see. I'm going to miss him.

I find myself staring at the navigation images, not making much progress, and decide I should go home and sleep. Surprisingly, I actually do this. On the way out I glance back at my chair and see that someone's posted a sign on the back: "How's My Driving? Call ...."




Footnotes:

[1] An abbreviated version of the story: on sol 41, the Spirit team initially planned a go-and-touch -- an unsafe maneuver that would drive and use the IDD afterward, without humans in the loop -- and they called Bob to check on it. This meant they woke him in the middle of his sleep cycle to ask him if they could do something he'd already nixed a dozen times; this time he got extra angry about it and let them know in no uncertain terms. Art Thompson started calling him "Angry Bob" after that, and the nickname stuck.

Now, the nickname wouldn't have stuck if Bob didn't also have the behavior described above, where he'd get increasingly exasperated as the night wore on and expressed that exasperation, shall we say, colorfully. Still, to be fair, the night they woke him to ask about the go-and-touch, Bob was right to chew them out. He'd tried everything else, and it hadn't worked. Chewing them out did work: we never tried a go-and-touch again, until a much later software upgrade made it safe.

For the record, "Angry Bob" prefers to be called "Concerned Bob." :-)

No comments: