2012-02-27

Opportunity Sol 1101 (Spirit Sol 1121)

It's a busy day. We start off with a MB soil-touch test, just planting the MB on the soil to see if the contact switches trip -- not that there's anything wrong with Opportunity's MB, but this exercises most of the same software path that's a possible culprit in the Spirit MB contact problem. Whether we see the same problem on Opportunity or not, we'll help narrow down our suspects.

After that, we're driving, putting some distance between us and Cabo Corrientes. The first part of the drive uses Visual Target Tracking, a new R9.2 flight software feature we're still checking out. After that, it's a cool 30m of autonav toward the Cape of Good Hope. You know, the one on Mars, not the original.

We need the help of the main VTT developer, Won Kim, to be sure we're doing today's part of the VTT checkout right. I am quickly reminded that research guys aren't used to ops -- they don't have the sense that you gotta do stuff and make decisions right now because there's a tactical deadline, and if you miss it, you blow a sol. So they aren't in the constant hurry that characterizes the rest of us. They dither. They optimize. They tweak.

They drive me nuts.

So I put him with Terry. Which isn't mean; it's a brilliant (ahem) solution because Terry himself comes from that world but understands this one, so he can speak Won's language and still get things done on time.

And it works out just that way. Despite some early delays, we're mindful of the hard 14:30 deadline that Matt Keuneke spells out for us, and we meet it with room to spare. Hell, we're ready at 14:00. It's almost embarrassing.

Since we have all this extra time, maybe we can afford to dither, and optimize, and tweak. Or, come to think of it, let's not.

[Next post: sol 1105 (Opportunity sol 1125), March 3.]

2012-02-22

Spirit Sol 1116

Today we're continuing to follow our months-old tracks back toward Home Plate. Our plan had originally been to skirt this little ridge about halfway along today's drive path. But then I went hunting for any old images we had of the ridge, to see what it looks like on the far side, just in case the far side hid a gimpy-rover trap.

It turns out we have a fantastic, close-up image of it -- as we were running for McCool, shortly before the RF wheel failed, we stopped right in front of it at the end of one drive. So we're able to look closely, decide that the slopes and pebbles are all benign, and commit to going over it rather than around.

That still leaves us with another decision: are we going to go north, along Home Plate, to image its east face? Or will we go west, toward the on-ramp? Or, maybe, will we compromise, heading west past the debris field and then heading north? We won't be doing that north-or-west drive today, but we need to set ourselves up so we can do it tomorrow. Since it's still a matter of some debate among the science team, however, we compromise, aiming instead for a point where we'll be able to choose either path on the next sol.

I used to not like compromise. But when it gives other people badly needed time to make up their minds, and when it makes your own life simpler in the process, well, there's something you can say for it.

[Next post: sol 1121 (Opportunity sol 1101), February 27.]

2012-02-20

Spirit Sol 1114

Damn, we nailed that drive. The rock's right in front of us, right smack dab where it's supposed to be, ready for us to give it a working-over with the IDD.

But will we? The IDD diagnostic we did on 1109 gave us troubling results -- no news, which in this case is bad news. We tried touching the MB to the soil, and we got nothing. The images and telemetry agree that there was simply no contact.

Now, that's weird. We really should have seen contact. This makes no sense.

Whether it makes sense or not, the consequence is that we can't rely on the MB to sense contact now.[1] That means a lot of IDD work has to get put on hold -- we can't place either the MB or the APXS, and we can't touch soil at all. It's RAT, MI, or nothing. Indeed, it's worse than that. Since we're not sure what the failure is, until we've done more analysis we're not sure it's safe to move the IDD at all.

The question for the science team is whether they're willing to wait here until the analysis is done, or move on. It's a painful choice, since we worked so hard to get here, but they decide to take pictures and move on. With stolen uplinks, upcoming restricted sols, and other issues, if we decide to stay here at all we'll be here for a week, and they -- by which I mean Ray -- are not willing to spend that much time for this observation.

He puts it well at the SOWG meeting. After a good summary of the IDD situation, he says, "The main reason for coming to Bellingshausen was to do remote sensing on Troll-like outcrops. We've done that. We should drive on and do continued IDD checkout as we go."

That argument carries the day, so we're on the road again. It's sad to leave Bellingshausen, since we were so proud of the drive, but that's the game.

One thing that made the drive so impressive was the damage it did to the terrain. With one wheel stuck, we chew up a lot of soil just turning around, and the previous turn was a 180. So the RHAZ from today was quite a sight, with a huge swath of Martian topsoil shoved around by our anchor.

So impressive is it that Squyres calls in to ask us about it. He's teaching the Intro to Planetary Astronomy and Exploration this semester at Cornell -- who better? -- a class of about 250 students. "Every day I give them about a 5-minute rundown of what the rovers have been doing," he says. "It's kind of a fun way to kick things off. Today I showed them that image of the churned-up soil and all the back-and-forthing from the downlink report."

So at least it'll be good for something.

Today is an RDO, meaning people who work JPL's newly introduced 9/80 schedule have the day off.[2] As a result, lots of Lab services are closed; in particular, only one of the cafeterias is open. Then, just a few minutes into lunchtime, there's an announcement: there's been a fire in the cafeteria. Nobody's hurt, but they have to close that cafeteria. Project Manager John Callas, however, buys everybody pizza from Round Table. Our hero!

I blame the disconcerting announcement of the fire, and my concern about the safety of the cafeteria personnel, for what happens later. For the first time since I don't know when, we catch a real error at the final walkthrough and have to redeliver. (Unusually, Sharon's here and watching. "It's probably my fault," she grins.) The problem was that our drive backbone sequence didn't check whether its helpers were on board, and if they failed to make it to the spacecraft, a variety of bad things might have happened. Annoyingly, I'd briefly thought of this problem earlier, but I failed to write it down, and it went back out of my head by the same door it came in, I guess.

So we have to walk through it all again, and when the time comes to do that, Rich Morris -- a Mission Manager and an RP wanna-be -- asks if he can do the walkthrough.

Ashitey and I glance at each other, then shrug, intrigued. "Go for it," I tell him.

He does it! And he does a pretty good job of it, too.[3]

[Next post: sol 1116, February 22.]





Courtesy NASA/JPL-Caltech. Can you believe that when we landed on Mars, I felt bad about every imprint I caused the wheels to leave in the soil? And now look what I'm doing to the place. You'd think I lived here.




Footnotes:

[1] This is because the MB instrument on the end of the arm has the only contact plates designed for sensing contact with the soil. The APXS and MI, in different ways, can sense contact with hard surfaces -- though the APXS's way of doing that has been broken for a long time anyway. But only the MB can sense soil contact.

[2] In case you don't have this at your workplace: 9/80 means you work 80 hours in not in the usual 10 days but in nine, by working an extra hour for eight of those days. Then you get a day off every two weeks -- in JPL's case, you get every other Friday off. Unlike most of the Lab, MER stayed on the standard, 5/40, schedule. I am personally on something more like 5/80. :-/

[3] So it's a shame Rich is no longer with us: he killed himself last year. This is the first time I've been able to see, up close and personal, the effects that a suicide has on the people who are left behind -- and I know it's much worse for his family than for his co-workers. Rich had a number of good friends, both on MER and off MER, who would gladly have lent him the perspective on life that perhaps he lacked. If you're in a similar situation, find a person in your own life who will give you that perspective, and make sure you get it from them. Please.

2012-02-15

Spirit Sol 1109

I'm happy, so happy. There's a MER science convention in town, and the science team is actually here in the room with us. Today we're joined by "PANCAM Emily" (as we call her, to distinguish her from "Mission Manager Emily"), along with a couple of other usually off-Lab science team members.

Others flow in and out through the day, including Rob Sullivan, whom it's always a great pleasure to talk to. And that's not only because he makes a point of profusely thanking the RPs for all of our hard work, and enthusiastically sharing the science results he's deriving from the observations we're helping perform. Okay, that doesn't exactly hurt, but mostly it's just that he's a super-nice guy.

Another side benefit of having the scientists in town is that they give a science update/briefing for the engineering team. Except, of course, for those of us on shift -- such as yours truly. It really sucks to miss it, but if you have to miss it for something, it might as well be this.

But there's one part I refuse to miss. Word reaches us that one of the scientists brought in an actual Mars rock -- a couple of them, really; ancient Martian meteorites that have been found here on Earth. The moment I hear about this, I zoom upstairs to get a look. Objectively, they're kind of disappointing, just a couple of chunks of granite-like stuff. But the idea that I'm holding a piece of another world in my hands is plain thrilling, even if I do have to wear latex gloves to do it.

To my astonishment, I literally have to drag Terry and Ashitey away from the keyboard to get them to go up and have a look themselves. ("Guys," I point out, "the tactical timeline is important, but this is a once-in-a-lifetime opportunity. Do not miss this. You will kick yourselves if you do.") As it is, Ashitey gets distracted by Jennifer Trosper and never gets around to holding the rocks -- as he sheepishly admits to me later -- but Terry makes a point of thanking me.

Despite all the distractions, we get through our sequencing without too many hiccups. And it's a complex day, with both IDD and driving: we're doing a test of the MB contact switches, then spinning (well, lurching) around so we can examine the Bellingshausen outcrop we dragged ourselves over to on a previous drive. The drive sequence is awfully complex, but we're getting better at these drives, forming tested and reusable pieces to build them out of, so that they're not quite as bad as they look.

Squyres is one of the visiting science team members, of course, and when he stops in for the final walkthrough, he can barely contain himself. "If you'd done a drive like this during the nominal mission, the Mission Manager woulda thrown you outta the room!" he cackles.

Since he's here, I take the opportunity to ask him about, well, Opportunity. (And about the HiRISE cameras, which seem to be showing some kind of premature degradation in the optics or electronics. He's glum about that.) The plan for Opportunity, Steve says, is tentatively to visit two more promontories, and then we go in.

"At Duck Bay?" I ask him.

"Maybe not. We can ingress wherever," he says. "At Endurance, we had the rule that we could only go in if we could prove we could get back out. Here we have no such restriction: the project is OK with it if we're unable to egress."

Well, the project might be OK with it, but I'm not. Now's not the time to make a point of it, but we're not going in there unless we can get out again.

Speaking of getting out ... it's Valentine's Day, and Sharon has decreed that we are not to remain here past 5 PM. She shows up about 20 minutes before that so she can start glaring at everyone a little early.[1] Despite a late start, a drive plus IDD on the same sol, and a multitude of distractions, we make it. Barely -- but we make it.

[Next post: sol 1114, February 20.]




Footnotes:

[1] Just one of the many things that made Sharon probably the best boss I've ever worked for.

2012-02-09

Spirit Sol 1103

Hot damn! She can still dance!

I can't sleep, of course, so I get up and check out the downlink. It takes me a couple of minutes to realize that the reason it looks funny is that she made it, and then some. The reason I don't see a big churn of soil right in front of us is that it's at the other end of a long drag mark, meters away now.

Not only did she turn completely around, but she also managed to complete nearly the entire bonus drive.

Baby, I'll never underestimate you again. That's a promise.

What do you do with a rover this feisty? You give her a bigger challenge. And that's what we're up to today. Continuing to follow her old tracks back toward Home Plate, we plot a course that slaloms backward -- northwest, west, northwest again -- for a total of 11 meters or so. Part of that, again, is bonus driving; we'll take it if we can get it, but we don't expect to ... wait a minute, what did I just say? I'm not doubting you, baby, I promise!

For all that, it's not a particularly complicated day. We've been developing a reusable library of sequences to help build chunks of the drive, and we're able to use them to build most of the drive today. This confirms that the Observer Curse continues. (The Observer Curse is this: when I bring in a new RSVP developer to watch the process, as I've done today, it always turns out to be a fairly simple day. It's good that this leaves us time to answer questions, but when it's not a pressure cooker, it's hard for an observer to appreciate the need for rapid, solid tools. Oh, well. I'll have to communicate that need to Bruce-John[1] some other way.)

And by the way, I learn at the SOWG meeting that today is the first day of Martian Spring. No wonder Spirit is in such a good mood! She's feeling her oats and rarin' to go.

And if we don't come in tomorrow and find that she's gone the whole 11 meters, I'll eat my hat.

[Next post: sol 1109, February 15.]





Courtesy NASA/JPL-Caltech. Looking back at the long right-front-wheel streak we left on this past drive.




Footnotes:

[1] John Baldwin was a developer we brought in for a while to work on my part of the RSVP software, RSVP-RoSE. Since we already had a John (John Wright), we called the new John "Bruce," or sometimes "Bruce-John." It's a reference to a Monty Python sketch. And I remind you that there is no Rule Six!

2012-02-08

Spirit Sol 1102

We didn't see the MB's contact switches trip on Puenta Arenas, after not seeing them trip on Mount Darwin a few days ago. This is getting a little worrisome -- it's not too surprising that we had a problem at Mount Darwin, since the target was in the air, but we sort of expected to see contact at Puenta Arenas.

It's remotely possible they're not working any more, which would be awful; the MB contact plate is our only way to reliably sense the soil, so that would take out our ability to IDD soil targets. However, the switches are redundant for this reason, and unless there's a pebble stuck behind the contact plate or something, everything is probably just fine. We'll have to do an experiment on this in the near future.

But not today. Today we're driving.

If you can call it that. We're going to spend most of the day just turning around -- a slow and painful endeavor on a vehicle with a broken wheel. We have to use the fact that the RF wheel is an anchor, by pushing and pulling the whole vehicle around it. It's not precise, and it's not pretty, but it's tested and it works.

And it's slow. Every turn segment -- a single maneuver gets us about 14 degrees or so -- needs a visodom update afterward, and that costs us around 3 minutes. Since we need to turn about 180 degrees, we need about 13 of these, more or less. We'll end up spending 40 minutes or so just turning, out of a total drive time of a little over an hour. Take away the overhead at the beginning and especially the end of every drive, and we'll be lucky if we make any progress at all.

But the right way to drive this vehicle, it seems to me, is to be optimistic. What happens if everything goes right? Well, we might possibly manage to haul ourselves another 3m or so toward where we actually want to go. So Khaled and I add in some commands that will take effect if we actually complete the turn, and we'll see how far she gets. As long as she gets mostly turned around, we'll be happy.

The poor girl. She's so much more agile than this, at least in my memory of her. She's like a dancer way past her prime -- arthritic and broken, but not reconciled to it. Not yet. Maybe never.

2012-02-07

Spirit Sol 1101

Brenda Franklin is back and feisty again, and she's made us hundred-sol cookies. Chocolate chip, this time, in addition to the yummy ginger cookies she normally makes every time the last two digits are "00." Of course, we're not just planning sol 1100 today, we're also planning 1101 -- sol 13 in binary, as I point out.[1] ("I hope we don't have a sol-16 problem," John Wright quips.)

Yestersol's IDD placement didn't make contact with the MB, which has us a little worried. John looks carefully at it and points out that the target was floating above the mesh, which probably explains it. But this really sucks, since it means we've probably blown a science observation and might need to redo it. I send email to Ray and hope for the best.

Working with John, with no shadow, is such a pleasure.[2] It's rare these days that I get to work with one of the old-school RPs with no shadows around, and it's just so damn nice. We're both on the same page, and everything goes smoothly. Just like it oughta.

And you can tell those old days are getting pretty old. Today, Opportunity's drive will take her over the 10km mark! Spirit has 7km coming up, but Opportunity is probably going to continue to increase the gap from now on.[3] We'll see if the whole slow-and-steady thing works out.

But there's one thing we two old-school RPs fail to notice: we forget to enable the MB_2 contact switch before doing the MB placement. Kind of a rookie mistake, actually; and normally, fr_check catches this, but today it doesn't because of an unusual sequencing choice. (Which I later fix it to catch.) Happily, Rich Morris was paying attention in the IDD Flight School briefings I gave at the IST team meetings, because he catches this at the walkthrough.

Maybe he should start shadowing as an RP. I think we might need him.




Footnotes:

[1] I make jokes in binary? Wow, I really am a geek!

[2] As you'll recall, John Wright and I were two of the original eight rover drivers -- and two of the four original Spirit drivers, at that. John is also one of my favorite people in this world, someone who makes me feel lucky to have the life I have. I'd never tell him that. :-) But it's true, all the same.

[3] Yep. Opportunity's now past 34km -- the second-most-traveled vehicle in the solar system, after Lunokhod 2. And we're going to break that Soviet record if it kills us.

2012-02-02

Spirit Sol 1097

Ugh. I'm sorry, but getting here at 07:30 is just brutal. It puts me in a bad mood.

And I'm not the only one. Ray Arvidson is in a bad mood, too. He's often kinda grumpy, but today he's particularly grumpy, snapping at engineers and fellow scientists alike. When there's some hesitation at the SOWG meeting about picking a name for thisol's IDD targets, he half-snaps, half-whines, "Just pick a name, folks, so we can move on."

Later, I propose offline that we should use a new target namespace -- the names of different kinds of bugs. Ants, grasshoppers, crickets, etc. And the goal would be to guess which kind of bug crawled up Ray's butt today.[1]

For our relatively simple two-sol IDD work today, Khaled -- shadowing, and so in the seat as RP-1 -- decides to use one of the new R9.2 software features, AutoPlace. This lets the IDD FSW itself figure out how to put the IDD on a target, so we have to do less work. The mode Khaled's using it in is a kind of hybrid mode called "manual autoplace."

"Is that anything like jumbo shrimp?" I ask him. But not in a grumpy way.

[Next post: sol 1101, February 7.]




Footnotes:

[1] I feel the need to be painfully clear about this, though: Ray is a great, great guy. On this particular day he was pretty grouchy and I wanted to make a joke about it -- maybe in part because I was pretty grouchy myself. But I can hold any one data point like that up to hundreds of others showing Ray in a terrific light. He's a great guy.

2012-02-01

Opportunity Sol 1075 (Spirit Sol 1096)

As Steve Squyres points out, "We're falling victim to the extraordinary scenery we see before us -- flash is extremely tight." And he's looking to make cuts, and of course, one of the cuts we end up making is to reduce the quality of the ultimate and penultimate drive images, just after Ashitey made a (justified) stink about it. Oh, well.

So we're driving, but it's not a big drive, just a few meters, to get into position to do another piece of our long-baseline imaging. But because of all the tricky sequencing now required to leave useful tracks for VO in this terrain, it ends up being quite a lengthy and complicated sequence.

So much so, indeed, that Steve can't resist tweaking us about it at the walkthrough. "This has gotta be the highest sequences-to-meters-driven ratio ever," he says. "This sequence is a work of art -- this has gotta be the fanciest three-meter drive I've ever seen. This is what happens when you give the rover planners one and a half hours to go three meters."

"Back at the beginning of the mission," he continues, "we'd have done this with a single three-meter arc."

"But then your stereo baseline wouldn't be as good," Terry points out.

"And we'd have fallen over the cliff, yeah," Steve replies. "But apart from that ...."