2009-09-24

Spirit Sol 259

Thisol's TUL is Julie Townsend, and she's a little distracted. She's trying to finalize her (re-)enrollment at Stanford, where she's getting her Ph.D. She's somehow finagled a deal where she'll basically do her regular job here at JPL and get a Ph.D. for it.

Damn, I didn't even know we could do that.

Today is the first time we're trying to plan three sols at once. It's a Friday, and the idea is that we'll plan the entire weekend today -- sols 259, 260, and 261 -- and return Monday. As far as the RPs are concerned, we have two sols to plan: on 259, we do a little IDD work, and on 261, we stow and bump backward uphill.

The uphill bump is the tricky part. We drove down to Conjunction Junction with a specific plan, part of a never-to-be-repeated (well, probably) wind experiment. Before conjunction, we imaged this area, then drove onto it, and the remaining part of the plan is to back out the way we came in and image it again.

As always with these vehicles, the hitch is that distance is easy but precision is hard. And this drive is all about precision. The images will be taken immediately following the drive, which means we have to choose the image pointing now, without knowing exactly where the rover will actually end up. To make life just a little more complicated, we need to get the rover back to the same spot as the pre-conjunction imaging, to make it easier to compare the images. Chris has already done the drive -- it's a simple one, just a straight backup -- but he hasn't allowed enough for slippage.

When coming downhill, we slipped farther than we'd expected -- about 45cm on a 1.5cm drive, giving 30% slip. I tell this to Chris, and he makes a face. "Thirty percent slip -- on this terrain?" he says.

"I was surprised, too," I shrug. "We were expecting about 10%, maybe 20% in the worst case. But both Rich Petras and I came up with the same number." I show him the result in the HAZCAM mesh in RSVP; when you model the rover driving 30% too far, the result in the front HAZCAM mesh looks almost exactly like the actual post-drive mesh.

Chris frowns again. "We shouldn't have slipped that much," he muses. He thinks for a minute. During conjunction, we got NAVCAM images from our new position. It should be possible to compare these with pre-drive NAVCAM images, which would give us a better slip estimate. Chris does this, and what do you know -- according to his results, we actually slipped only about 7%.

Good rover. Good girl.

But what the heck is up with that? Rich Petras isn't around, so we can't ask him how he did his analysis, but we realize that he must have based it on the HAZCAMs, since that's all we had at the time. The RSVP-based analysis I did is also ultimately based on the HAZCAMs, for the same reason. And since the HAZCAMs give less precise ranges, we postulate that the HAZCAMs just misled us.

We might be doing the wrong thing, but the smart gamble is to trust the NAVCAMs. So we do that.

Not for the first time, I marvel over the amount of blood, sweat, and tears that we sometimes shed over a matter of centimeters. Here, we sweated for an hour over a matter of maybe 20cm or so -- less than 8 inches. But when that 8 inches could determine whether we manage to nail a one-of-a-kind scientific experiment, well ....

So Chris's skepticism probably saved the experiment. But I have him beat: he saved the day only once, but I manage to save the day twice. I am on the ball.

The first one is definitely a case of Murphy's Law in action. Early in the mission, I wrote a flight-rule checking script, called fr_check, to help us RPs avoid dumb mistakes. One of the rules it enforces is this: we're supposed to open the MI's dust cover exactly when we're going to take an image. We should open the dust cover, take an image -- or a series of them -- and then close the dust cover again. On rare occasions, we deliberately take images with the dust cover closed.[1] (The dust cover is not opaque but a translucent orangish material; cover-closed images are poorer quality, but are helpful in making false-color products.)

Since fr_check can't tell whether we're taking a cover-closed image on purpose, this rule turned out to be a little "chatty," too eager to warn about cases where in fact everything was OK. Warning too frequently can be just as dangerous as not warning at all, since it can lead you to ignore actual problems, so one of the other RPs recently asked me to dial it down a little.

And sure enough, thisol -- the first MI-using sol after I made the change -- the change caused fr_check to fail to warn about an actual problem. Thisol's MI work is a little unusual -- normally, we take a series of images at different distances from the target, but because we're so downlink-constrained thisol, we just took a single image. And we forgot to open the dust cover. Since there was just one image, the amount of MI work fell under the rule's new threshold, and it decided to keep its mouth shut.

Luckily, I catch this manually and add the appropriate commands to open and close the dust cover. (It ends up returning an absolutely perfect image.)

I save another science observation as well. Originally, it was our practice to call instrument-on and instrument-off sequences in the RP sequences. For instance, we'd place an instrument such as the MB and call the sequence that turned it on; later, we'd turn the first instrument off when switching to another instrument. For various reasons, we don't consistently do this any more, so it's reasonably common to have an IDD sequence that doesn't turn on the instrument it places. Instead, the master sequence -- the day's backbone -- will start or stop the instrument at the appropriate time.[2]

But today, after reviewing the master and IDD sequences, I realize something is missing. "Hey, did I miss it," I ask, "or did we not turn on the Moessbauer?"

Nope. We did not.

"That would have been embarrassing," I say wryly. One trivial change later, and I've saved a twelve-hour science experiment.

So even though planning three sols keeps us here late, I'm in a good mood. I always am, when I know I've earned the paycheck.

[Next post: sol 265, October 1.]




Footnotes:

[1] Sadly, on both vehicles, the dust cover has now accumulated so much dust that it is opaque. Better the dust cover than the lens itself, though.

[2] Now we always do it that way. It's easier to make sure you're getting something right when you do it the same way all the time.

2009-09-23

Spirit Sol 258

Mars is once more visible from Earth. And we rover planners are back to work.

Thisol's sequence is mostly cut and pasted from previous sols: we're once again taking MI images of the capture and filter magnets (though not as many as we usually do when we do this, since we're downlink-limited thisol), taking another image of the dirt clod stuck to the MB, redoing the MI mosaic over GreenEyes, and finally APXSing the hole we left behind when we pulled up the dirt clod.

The only thing complicated about the sol is putting the APXS on the dirt clod's former home. Normally, we have a terrain mesh that shows our targets, but that's not the case here -- this time, we have only the MI images to go on.

This ought to be easy: we know where we placed the IDD previously, and we can see exactly where the missing soil is relative to that placement. (The MB contact ring, which made the impression, is an O-shape. The resulting impression is a rough semicircle, centered "northwest" of the center of the images). The problem is that I'm not sure of the dimensions of the contact ring, so although I know which direction I want to move the IDD, I'm not sure how far to move it. The documentation I normally use to answer questions like this doesn't tell me, so I start looking around for someone who would know.

This proves to be harder than I expected. The Mobility/IDD guys on shift don't know. Jake Matijevic points me to a book of images, which shows the part I'm interested in but doesn't show scale. Lori Shirashi isn't around, so I call Joe Melko, who doesn't provide an answer.

At last I try asking thisol's MB instrument rep, Iris. She proves to know just enough: she knows the inner diameter of the O. Armed with this, I'm able to go back to the image Jake pointed me to, measure carefully with the ruler built into my trusty Swiss Army knife, and work out the width of the ring itself. (For the record: the inner diameter is 15mm, the ring itself is 10mm wide.)

Which means, since I want to move the arm to the middle of the ring's impression, that I need to move it: 12mm. Thank you, and good night.





Courtesy NASA/JPL-Caltech. An MI of the filter magnet, showing -- yup -- we plopped a chunk of dirt right onto it.


Courtesy NASA/JPL-Caltech. The capture magnet, which we also imaged thisol. I love that manadala-like bullseye pattern.

2009-09-06

Spirit Sol 241

If all goes well, I don't have a job. For the next ten days or so, that is -- we won't be driving the rover or moving the IDD during conjunction, so I can spend the time on other projects.

If all goes well. Which, again, we won't know until toward the end of the SOWG. But the plan is that there won't be anything for me to do -- and, indeed, the confirming NAVCAMs of the last IDD placement arrive on schedule, and the IDD is right where it's supposed to be. I flip back and forth through the images, and everything just looks like it went perfectly.

Luckily, there are sharper eyes than mine looking at these images. Rich Petras spots a dark clump on the MB, one that's not supposed to be there. When we look more closely at the images, we discover that the MB picked up some dirt when we used it to touch the soil. (The MIs also clearly show this dirt missing from the soil patch itself.) From that point on, each image that shows the MB shows the chunk of dirt riding along with it, hanging off the outside of the contact ring.

Including -- uh-oh. Including the image we took just before placing the MB on the filter magnet. Which means we just put a whole bunch of dirt onto the magnet.

It's unlikely this will be a major problem. Probably, the worst case would be that when we placed the MB on the magnet, the closing contact ring squished the dirt into the MB's own field of view. If indeed this happened, they'll likely be able to recalibrate the MB to subtract out the dirt's effect. And maybe we'll find a way to shake it off, or scrape it off, if it doesn't look like it'll fall off by itself.

I ask the scientists whether they want us to do anything about this nextersol -- maybe we could retract the MB, image it to see if the soil's still there, check out the magnets, whatever. But they decide against it. I've put the arm where they want it to be right now, and they don't want to risk that. They'll want to inspect the MB and the filter magnet more thoroughly, but they want to wait until after conjunction to do it.

So that means no IDD work nextersol -- nor, indeed, through conjunction -- and therefore nothing for me to do.

Well, I guess there is one thing to do -- we need to start laying out a long-term post-conjunction drive plan. The next science goal is to reach Pergamon Ridge, about 75m east from Conjunction Junction. It's pretty easy to plan this path; there's no good way to get there from here, but Ray and Bethany and I look at the insolation maps and decide we'll head to the northeast. As we go, we'll keep taking pictures of Pergamon Ridge. If a path opens up, we'll take it; otherwise, we'll have to treat Pergamon Ridge as unreachable.

And that's it. I wish everyone a Happy Conjunction, and go home.

[Next post: sol 258, September 23. Sorry about the long delay; that's what happens when a gigantic ongoing multi-billion-year-old nuclear explosion floats between you and something you want to send radio signals to.

Most of us spent these two weeks catching up on other work, or taking vacations. I did some of the former, but used about half the time to write an ops tool that had been sitting in my head for months.

I called it "mPhoto," and it was conceived as a sort of "iPhoto for MER." The idea was to quickly render a simple grid showing thumbnails of all of the recent images, enabling the user to rapidly zero in on images of interest. In addition, it provided some very simple image-manipulation facilities, and the ability to save the images in more useful formats than the one the project normally uses internally. Partly, I wanted the tool itself, and partly I wanted a chance to experiment with writing software in a different -- frankly, a somewhat sloppier -- way than I normally do.

It didn't come out perfect, but I was very happy with it, given that I explicitly gave myself a time limit of one week to develop the entire thing. Despite its limitations and lack of polish, it still does some things better than any other image-browsing tool on the project, so that even now it's still used by me and others. It scratched my itch, and I had fun making it.]






Courtesy NASA/JPL-Caltech. One of the NAVCAM images showing the chunk of dirt that stuck to the MB, shortly before the sequence would put the MB on the rover's magnets. You can see the chunk of dirt on the MB contact ring, near the top of the image.





Courtesy NASA/JPL-Caltech. Sure enough, the MI images -- like this one -- show a chunk of dirt got torn out of the soil where the MB contacted it.

2009-09-05

Spirit Sol 240

Our email problems and the attendant scheduling problems continue, but it seems to be getting better. I arrive an hour early -- thinking I'd be only half an hour early, but the schedule started later than I thought. No problem. Early is no problem.

The day is ambitious. In the morning we MB-touch the soil and start a short APXS integration. After that, the fun begins: we'll do a 2x2 mosaic of MI stacks around the chosen target, hold the RAT out where the PANCAM can image its magnets, then MI the capture and filter magnets (which are located on the rover body, at the base of the mast). Last, we'll place the MB on the filter magnet.

As on previous sols, we're planning in the blind here; the downlink doesn't start until the SOWG meeting is nearly over. When we get the images, I'm at first satisfied -- then concerned. (Then ... kind of sleepy ... no, wait. Leave it at concerned.) Something looks wrong with the final front HAZCAM, and after a minute I realize what's bothering me. It's too level.

We wanted to end this drive with a better insolation (solar tilt). Stopped on the slope, with the solar panels aimed at the sun, would be good. Level is bad. It's hard to tell, but the picture looks level.

Oh, shit.

This means we're going to have to scrap the planned IDD work for the sol and hope we can work out a backward drive upslope -- and hope like hell that that drive works.

But before I panic, I check the image headers and do a little math. Our roll is -11.6 degrees, pitch -7.0 degrees, making our overall tilt about 13.5 degrees. RSVP's predictions varied depending on which terrain mesh you used, but it tended to show more like 20 degrees. So this looks bad.

What the hell happened? I ignore the remainder of the meeting, load yestersol's sequence into RSVP and simulate it. Like yestersol, I look at 10% slip (15cm) and 20% slip (30cm). Then I do something I should have done yestersol: I consider worse slips. At 33% slip (45cm) the predicted HAZCAM in RSVP comes out looking almost identical to the actual we just received.

So now we know what went wrong. It was exactly what I was worrying about during the CAM: we slipped more than expected, to the point where we started to level out.

I think hard for a minute about how I want to handle this. As the SOWG meeting breaks up, I activate my mike and tell Ray we might have a problem.

"We don't know that we have one," I tell him, "we'll have to do some more analysis first. But I wanted to let you know we might have to replan the sol to drive instead."

To my surprise, Ray sounds unconcerned. "Okay," he says calmly, "you guys figure it out and we'll do what we need to do."

First I go talk to Rich Petras, the Mobility/IDD guy on duty today. He agrees with me about the slip results (foo!) but doesn't know how to compute the insolation factor from tilt. Neither does our ACS guy, though he's able to show me what the rover will look like as seen from the sun. It looks good, good enough to start me thinking that we might be able to stay put after all. Then again, I don't have a sense of how good it needed to look. On a scale of 0.0 to 1.0, we needed to be about 0.88. Yestersol, we were at about 0.91; if we worsened at all, the rover will survive but we'll feel like dumbasses (that is, I'll feel like a dumbass). We were hoping for something like 0.94 or 0.95. And I can't visually distinguish an attitude that would lead to 0.88 insolation from one that would lead to 0.95 insolation. We need math.

Chris is thisol's floater, and he's in the sequencing MSA when I

return.
"I think we might not need to drive, so I want to get started on the IDD stuff," I tell him. "Can you figure out what our insolation is?"

"Already working on it," he says. He's slinging Perl to get us an approximation.

So he works on that, and I work on the IDD stuff. The script, when he finishes it, reports about a 0.89 or 0.90 insolation, which would be a dropoff from yestersol but good enough that we don't have to move. But only a minute or so after he finishes, Leo Bister comes in with the official word: we're over 0.93. Not only are we good enough, we improved. Not as much of an improvement as we'd hoped, but an improvement all the same.

There's also good news about my IDD sequencing. It's all going smoothly, almost suspiciously smoothly. The scientists have picked a their soil target -- Uxmal, a nice spot right in front of us. When I plug Uxmal into the sequence and run a simulation in RSVP, it works the first time. This is uncommon enough that I go back and check it, but, yes, it really is going to work.

I'm more or less done with the sequence and starting to do the detail work, when Bethany comes in with a request. They've just taken a close look at the PANCAM and there's a perfect patch of soil just by the right front wheel, which is a region that's generally awkward to IDD in. (As it happens, this spot would have been smack in front of us -- in ideal IDD territory -- if we hadn't slipped.)

"Can you get to it?" she asks.

Well, what the hell, I'm game. Everything else seems to be going my way today. She reads off the new target's coordinates and surface normal; I tell RSVP about it, tell the sequence to go there instead of to Uxmal, and simulate it. I'm all prepared for RSVP to spew errors that will take me hours of work to clean up, but nothing doing: like Uxmal, the new target works perfectly the first time.

I'm on a roll, baby.

I have a lot of little details left to attend to, and I even find time for polishing the sequences some. I love it when I find a way to shave a few seconds out of the sequence, to make it just that little bit more elegant -- not only for its own sake, but also because every second we save, every Watt-minute we can squeeze out of the engineering side, we can spend on science. And that's what we're here for, after all.

One of the reasons I have all this time is that, well, let's face it, I'm awesome. But another reason is that there are problems elsewhere in the plan that, like yestersol, slow us down significantly. Once again, we run hours late. Kevin says, "I thought yestersol was the suckiest sol ever. I was wrong. This sol is the suckiest sol ever."

I'm okay with it. My stuff's long since done and I'm ready for conjunction, but more importantly -- much more importantly -- I got a chance to do something I'd been hoping for since the beginning of the mission.

Most of the targets and features we encounter are named by the science team. Every once in a while, the rover drivers get to name something -- I've gotten to do it once or twice -- but it's not common. It's very cool to name something on another planet. The names aren't official, but they do tend to stick; and as when we RAT something, there's a sense of permanence to it.

Since the beginning of the mission, I've been hoping to name something in honor of Jake, since he died only a couple of months before we landed. It didn't seem like the right time when I had the other chances. This one felt right. I don't know why; maybe it's because I've been thinking about him lately, since we moved. Maybe it's some other reason.

Anyway, it fell to me to name the target we abandoned Uxmal for. The tiny little patch of soil our rover will be protecting from the wind throughout the long night of conjunction has a name: "GreenEyes." I miss you, buddy.[1]





Courtesy NASA/JPL-Caltech. The front HAZCAM, showing terrain that's more level than we'd hoped. Aw, crud.




Footnotes:

[1] I still do.

2009-09-04

Spirit Sol 239

The project's in a bit of a tizzy. The email server that handles all of the MER ops mailing lists was cracked a couple of days back, and as the system administrators scramble to get it back online, email service has become woefully unreliable. We rely on email to communicate meeting schedules, and the science teams use it to coordinate their planning, and nothing's getting through to anyone reliably.

So it's a bit of a minor miracle that the SOWG takes place at all, but take place it does. We still haven't gotten the downlink, so we have no idea whether yestersol's crucial drive went as planned. Consequently, we're planning in the blind -- and what a plan they come up with. Yestersol I left with the impression that -- assuming the drive worked -- we'd maybe do a bump drive if need be, and then sit still for conjunction. Somehow that plan flew out the window (or maybe I just didn't get the email), and they've decided to stuff in a bunch of complex IDD work before the drive.

I decide to take this as a challenge, and (despite my irrational but unshakable misgivings about the drive's chances) start sequencing the IDD work as fast as I can, while the SOWG meeting is still going on. I get help from an unexpected source: balky communication equipment. The science teams are mostly remote now, so the project invested in some fiendishly sophisticated hardware and software to enable us to interact. When it's all working, the remote teams are audible through the conference phones and visible on the big projection screens.

When it's all working. Which it mostly isn't. The system fails every few minutes, causing Brenda Franklin, the unlucky scientist stuck with running this stuff, to scramble madly around the room, apologizing to people, whacking buttons, and whispering urgently on her cell phone.

She's not happy. Neither is Ray Arvidson, who's annoyed even when the system is working at its best, because of the audio delays. There's a second or two of lag in the audio, just enough to make conversation difficult. One end of the conversation will sound like this: "I think -- I think we should -- sorry, go ahead -- okay, I think we should -- sorry, go ahead ...." The other interlocutor is saying much the same thing, but offset by a second or so, so they spend more time negotiating who should talk next than they do talking. (I think up a simple protocol: when there's a collision, the person who should proceed is the one whose name is alphabetically first. Arvidson would like this, but I doubt we can sell it to Squyres.) It's like when the anchorman talks to a reporter via satellite phone -- or, more apropos, like when Mission Control talks to an astronaut near the moon. And Houston, we have a problem.

Well, the good news is that I get all of the IDD sequences built even before the first bit of the downlink arrives -- all I need is for the scientists to pick a target, and I can plug that into the sequence and I'm done.

When the downlink starts flowing, I promptly pull up the final front HAZCAM and check out the coordinates in the image header. They're (0, 0, 0) -- we didn't move. Oh, fantastic. We are so boned. I knew I'd fuck something up.

Oh, wait. It's (0, 0, 0) in the new site. And, duh, the terrain is different. Holy shit, we did move.

By the time I've figured this out, the penultimate HAZCAM is down. This one will actually be more useful to me, since it contains the rover's estimate of its position near the end of the drive. That'll give me an idea of whether we got anywhere close to the destination. It's a trivial bit of math to plug in the numbers and -- we're about 1.25m from the destination. Which is excellent -- yestersol, we estimated we'd take this HAZCAM from about 1m or 2m away from the target, depending on slip, so we're right in line with expectations. Unless the rover wigged out in the last couple of drive steps, we should have made it.

When Jeng works it out more precisely, it turns out that we hit the bullseye. Yestersol's goal was to put the rover within 70cm of Tikal -- that is, we had a bullseye with a 70cm radius. Our actual distance to Tikal after this drive: 68cm. Jeng says it's the best drive he's seen in the last month, which might be unfair to the other rover drivers but makes me feel great anyhow.

Not much more data has come down, and a couple of minutes later we learn why. We screwed up the data volume predicts yestersol. It turns out our final attitude is very poor for getting data back for this comm pass (we estimated we'd get 55 Mbits, but we got only about 15), so we're going to have to plan everything using the HAZCAMs and not much else. In particular, the post-drive NAVCAMs aren't coming down. But we probably don't need them. We have acceptable NAVCAM coverage of this area from the previous sol -- we didn't drive that far, after all -- and the penultimate HAZCAM shows us the near field in good detail. We were planning to drive only a meter or so thisol anyhow, and I think we can do it with what we have.

To make matters worse -- or better, depending on your point of view -- part of the right-eye image of the final front HAZCAM was lost in transit due to the poor quality of the comm pass. This leaves us without stereo data for the region immediately in front of the rover, which means we can't plan IDD work for the sol -- without knowing exactly where the ground is, we can't safely move the arm. Consequently, all the work I did building the IDD sequences was a total waste -- but the upside of that is, I can focus all of my energy and attention on the drive, which remains in the plan.

The goal of this drive is twofold. The first goal is simple: if possible, further improve our energy position -- the more we can tilt the solar panels to face the sun, the better. The second goal results from some disputed observations from early in the mission. Some of the scientists think that when we were inadvertently stuck at Adirondack, some interesting things happened to the soil under the rover. (I'm not clear on what these interesting things were, but I think they had to do with the effect on the soil of the way the rover disturbs the wind.)

So the scientists want to use conjunction to do a soil experiment: before conjunction, we image a patch of dirt and drive onto it, disturbing it as little as possible with the wheels. (The interesting stuff will be the soil under the rover body; we just have to be careful not to turn in place or do anything else with the wheels that would disturb that part.) After conjunction, we back out the way we came in, and see what happened.

The resulting drive is, shall we say, elegantly simple. We'll turn 30 degrees to the left -- causing us to face down a short north-facing slope, so that damn near anywhere we stop on the slope will be a good solar attitude -- take some pictures, and drive a meter and a half. (It ends up being slightly more complex than that, as we need to pause halfway through to acquire some MTES images as well.) We get to associate a short name with every sequence; since this is the bump-drive that will take us to Conjunction Junction, I name this one "Conjunction Bumpin'." Then I sit and wait ... and wait ... and wait ....

Because meanwhile, life has been sucking for the rest of the uplink team. We have some unexpected heating requirements, which end up blowing the sol's energy budget and forcing Kevin and the TAP/SIE to replan the sol. (Happily, the drive is untouched, so I can ignore all this except for the delays it introduces.) In the end, they settle for just blowing the damned energy budget; they'll end the day with a net loss in the batteries, but they'll still be above critical levels. They'll make up for it nextersol. This all leads Kevin to declare thisol "The Suckiest Sol Ever."

Kevin's the type to muse aloud about his setbacks. "I think I was too nice today," he says. "I should have said at the SOWG, 'This plan is ridiculous, go change it' .... Tomorrow, I'm gonna be a dick!"

It's so important to have a plan of action.

So the CAM starts three hours late. This exposes yet another problem with remote science teams -- they're all in later time zones than ours. Ray's on Central Time, for instance, so our 22:30 CAM starts at 00:30 for him. He sounds snippy and pissed, not that I blame him.[1]

It's not the kind of sol where you want to find a problem at the CAM. Mindful of this, I keep my voice down when I point out a problem to John (who's floating again thisol). The slope we're heading down is a short one; if we slip too much, we'll reach the bottom and level out, leaving the solar panels aimed straight up, not at the (northerly) sun. This will make things worse for us -- much worse.

I roll my chair over to him. "You think we're going to have a problem if we slip too far?" I ask him quietly.

"I wouldn't worry about it," says John.

"OK, that's what I wanted to hear," I say, and I roll my chair away. I think about it. No, that won't do. I roll my chair back. "OK, why wouldn't you worry about it?"

"If we start to level out," he points out, "we'll stop slipping."

Duh. I roll my chair away again.

And I think about that answer. No, that won't do, either; by the time we've leveled out enough to stop slipping, our solar attitude will already be terrible. I need to know this, or I'm not going to be able to sleep, so I model 10% and 20% slips in RSVP. The results have to be taken with a grain of salt since our terrain data is so limited thisol, but ... 10% slip is fine; 20% slip isn't as good as the no-slip case, but it still leaves us with a good solar attitude.

And that's what I wanted to hear.




Footnotes:

[1] As I've learned since then, Ray is a classic morning person. He hates late hours, and like most morning people, he really hates Mars time. That didn't stop him from living on Mars time as the science team lead for Phoenix a few years after this, though.

2009-09-03

Spirit Sol 238

I'm running late and there's no parking, but I still manage to make it up to the SOWG room right on time. Where I find an ongoing meeting. Turns out they revised the schedule, bumping up the meeting half an hour. So I'm not barely on time, I'm horribly late. That's always a great way to start your day.

It also turns out they've changed the plan, again without telling me. (Though to be fair, maybe I'd have known this earlier if I'd known when the SOWG meeting was and I'd been able to get there on time.) We're going to make up for our lost sol by compressing the two drive sols into one: we'll drive to Anasazi_station, image Anasazi, and drive to Tikal, all in the same sol. But I don't learn this until after the SOWG meeting, so the duration estimates we use for the plan are way too short, which means we have to replan later.

I complain about all this to John (thisol's floating RP), and he just grins and says, "This is so cool ...." Once again, this simple phrase snaps me back into a good mood. Bastard!

So we need another thirty minutes for the extra driving (it's only about three meters and some turning, but visodom slows things down a lot), and this simply breaks the plan -- there's not enough energy to do it all. The science team decides to nix the dogleg to Anasazi_station, as well as the associated PANCAM imaging. (Thus disappointing Heather, who's already built the PANCAM sequences. But as always, she's a good sport about it.) This leaves us with a straight drive to Tikal and plenty of time to do it.

The more complex drive was already built, but I decide to follow Heather's example and be a good sport about it. Wincing internally, I rip out about three-quarters of the drive sequence and redo it as a straight drive to Tikal. The rest of the work is mainly analysis, to see whether we're going to be at a good solar tilt if we error out, and stuff like that. Turns out there's nothing to worry about: the insolation maps -- colored terrain overlays showing the goodness of our sun-relative tilt at each point in an image -- show the path from us to Tikal as being completely safe. In these maps, blue is good, green is so-so, red is bad. And the path from here to Tikal is a big blue highway.

Though still under the weather, Mark Adler is back as thisol's Mission Manager. Every Mission Manager has his own style. Mark's is to ask a few really tough questions to see if you seem to have your shit together. If you pass the test, his comfort level increases and he leaves you alone.

Thisol he decides to make the test open-book: he writes the questions down on the MERboard, and announces that this is what he's going to be asking at the CAM. Nearly all of them are for me, to wit: (1) What tilt limit was chosen for this drive? (2) How was that limit derived? (3) If the drive faults out due to a tilt error (or any other error), how will we recover in time for conjunction? (4) What is our insolation along the drive path?

Having the list in advance makes my life about a zillion times easier. I make time before the CAM to discuss my answers with Mark. (Which is the way to do it -- early, I mean.) As I'm talking to him, I watch the comfort level rise above his critical threshold. He visibly relaxes, and the drive is on.

I'm glad I've already gotten this out of the way when, ten minutes or so before the CAM, John and I notice a rock-ringed depression just in front of the rover and along our drive path. (Indeed, the encircled depression looks uncannily like a fire pit.) To a first approximation, it's not a drive obstacle. It's only about 20cm deep, from the top of the tallest rock to the lowest spot we can find. But there are some aspects of its shape that worry us. For one thing, the path from the bottom of the depression to the top of that tallest rock is nearly vertical -- a mini-cliff, which the rover would have a hard time scaling. Moreover, the area immediately in front of the mini-cliff is just about one wheel diameter across. This reminds me of a "feature" of the rover, a problem that arises because it can't lift its own weight. You can build a kind of rover trap, a hole just big enough for one wheel, which the rover will have a hard time extricating itself from. Put all six wheels in such a trap, and the rover turns into a lander.

We soon deduce that if the rover gets stuck in the fire pit, we'll be able to get it back out -- eventually. Probably, we could just have it turn in place and it would climb out the side. The problem is that we don't have any sols we can afford to spend doing this; if we get stuck in the fire pit, we may have to remain stuck until after conjunction. That's assuming that the attitude change induced by dropping a wheel into the pit doesn't blow our insolation, causing the rover to drain its batteries during conjunction and freeze to death. In which case, we'll still be there after conjunction ... and for a long time afterward. (OK, so that's a worst-case scenario. But worrying about worst-case scenarios is what they pay me for.)

Ultimately, we conclude there's no reasonable risk. The depression is quite near us, only about a meter away, so we won't accumulate much uncertainty before we reach it. The likely case is that we'll just drive straight over it, mostly straddling the depression, and in particular straddling the worst-case area.

Ah, it's always something.





Courtesy NASA/JPL-Caltech. The rock-ringed depression we started to worry about is straight up the middle of this image.

2009-09-02

Spirit Sol 237

I have an interview this morning, with a reporter from the London Sunday Times. He's here to do a car-review-style piece on the rovers. The sample piece he emailed me makes it clear this is to be a light-hearted article, and I'm looking forward to it.

Only one problem: the press office sent me a time for the interview, but not a place. We eventually get that sorted out; it turns out the reporter's running late anyway, giving me time to attend a meeting to decide whether we should drive before the upcoming solar conjunction period.

My feeling is, no: we're receiving a good amount of solar energy right where we are, so why risk moving? Better to be conservative; sit still for the few remaining pre-conjunction sols. A bird in the hand, etc.

But I don't bother voicing this view, as I don't feel strongly about it and it rapidly becomes clear that everyone else thinks we should drive. The scientists are keen to visit Tikal, a spot about 9m away. Nine meters used to be a doddle -- not so long ago, we commonly drove ten times that far in a typical sol. But driving on slopes is trickier; we've been having a lot of trouble with our drives since we started climbing West Spur. What's more, we might have a narrow target to hit, and precision driving with these vehicles is always hard, even on level ground.

The tentative plan that emerges is to spend sol 237 finishing up our IDD work at Ebenezer and driving eight or nine meters to a spot called Anasazi_station, where we'll take nice PANCAMs of an interesting-looking rock called, you guessed it, Anasazi. On sol 238, we drive from Anasazi_station to Tikal, only about another three meters. On sol 239, we do fine positioning if needed to tweak our attitude so we can get more solar energy. That leaves us a contingency sol before the hard limit after sol 240 -- due to the upcoming solar conjunction, we will not drive after sol 240 no matter what.

That plan's fine by everyone in the room. But Jim Erickson is not in the room, and as Art points out, Jim already nixed the drive to Anasazi_station -- Jim wants to keep things simple, and to get us to our solar conjunction destination sooner rather than later. (Which is pretty much what I was thinking. Evidently, I'm project manager material.) This surprises Squyres, and he and Mark Adler decide they'll try to change Jim's mind: apparently they've refined their estimate of the start of conjunction, and Jim might have been thinking we had a sol less than we do.

Oh, and incidentally, everyone has already picked out a name for the spot where we spend solar conjunction. Wherever that may turn out to be, we're calling it "Conjunction Junction."

You'll never guess what they're planning to use as the wakeup song.

That's the end of the meeting and the start of my interview. The reporter, John Arlidge, is a tall skinny fellow accompanied by a photographer (Paul Harris). "John and Paul," I say, shaking hands. I could resist if they weren't British, but as it is ... "Where are George and Ringo?" I ask. They have the good grace to laugh at this, and we're off to the ISIL so they can get pictures of me pretending to drive the testbed rover.

John decides to start the interview while we're waiting for Paul to set up his equipment. "Noisy in here," he says. "Let's try a sound check." He puts the tape recorder on a table next to me. "Say anything," he says. "Tell me what you had for breakfast."

"I didn't have breakfast," I tell him. "I was so excited to come here and talk to you, I skipped it entirely."

He laughs and gives me the finger. "Fuck you," he says cheerfully.

So this is a reporter I can deal with.

I give John a little background about the rovers -- their basic capabilities, environmental conditions, what we've been up to, where we're going, stuff like that -- until Paul's ready. Then I don one of the anti-static coats -- we really wear these things when working in the testbed, but I feel like a dope because I realize it's going to make me look like a pharmacist in a commercial for headache pills -- and walk across the Mars-like sand to the testbed rover. I grin idiotically at Paul while he snaps some pictures of me standing there, sitting there pretending to drive the rover with a laptop, and so on. His camera's digital, so he shows me some of the pictures afterward, and I tell him, "I think the ones where I don't look like a homicidal maniac are the best ones." He agrees.

So then we all walk back to the sequencing room in Building 264. John, being smarter than Paul and me put together, falls behind us as he keeps pace with Charli, the attractive young woman from JPL's media office. (It's her second time this morning walking uphill across the Lab -- in high heels -- and she's slowing down.) Paul, it turns out, lives here now, in Woodland Hills, but he gets back to London a couple of times a year to visit his parents. I ask him about London and he asks me about the rovers, and I get the definite impression he's more interested in the vehicles than John is. John is more interested in Charli. As I said, he's clearly the smart one.

We spend another 45 minutes or an hour talking in the sequencing room, though it seems less of an interview and more of a brainstorming session -- brainstorming about what funny and clever things John could write in his article. Whether there's valet service on Mars, what kinds of girls you could pick up with a Mars rover ("smart, science-oriented, and very patient ones," is my answer), that kind of thing. Meanwhile Paul takes pictures; I've set up a drive animation to loop on the projection screen behind me, and I try not to look into the camera as he flits and shoots. Paul eventually puts his camera away, but has to haul it back out when John points out the "How's My Driving?" sign taped to the back of my chair.

Then it's over. Or nearly. "My last interview before you was with Damon Dash, the producer behind Jay-Z," John says. "He got a free Ford Explorer just for driving it around and saying whether it was bling. I bet you could do something like that."

"I don't think our ethics people would like that," I reply tactfully. Not that I haven't thought about it. But really: the job is worth more to me than some lousy car. I already get to drive the coolest vehicle in the solar system, what the hell do I need with a monstrous, gas-guzzling Ford Explorer?

Meanwhile, back on Earth, the news is not so good. Yestersol was supposed to be a touch-and-go, but when we get the data, it turns out we didn't go anywhere. Apparently, Chris badly underestimated the IDD sequence duration, so when the rover tried to drive, it noticed that the mobility time-of-day limit had passed and refused to move. This jeopardizes the long-term schedule we worked out earlier in the day, because it means we'll have to spend thisol redoing part of yestersol -- effectively, we blew our contingency sol before we even started.

"So Chris is fired," says Adler. "But he's rehired, because we need him for today."

It's uncommon for us to redo the previous sol like this, but it does make the sequencing easy -- we can usually just pick up yestersol's sequence, make a few minor changes (or not), and we're all done. Which is how it works out today. I have very little real work to do, I mainly struggle heroically to catch up with the email backlog that accumulated while I was out of town. (The backlog is exacerbated by worms that automatically spew email across the Lab. I read my email under a niche mailer on a Linux box, so I'm immune to the worms themselves, but I still have to put up with the junk messages they send out in their efforts to propagate. My productivity would double if the Lab would ban Windows, and I don't even use the fucking thing.)

Adler's under the weather, so he goes home and Jim Erickson fills in as Mission Manager for the sol. Jim hasn't worked on Spirit in a long time, and he's alarmed by some things we take for granted, like driving on sharp tilts. We're expecting a max tilt a little over 20 degrees thisol, so I set the limit (beyond which the rover will automatically stop moving) at 30 degrees. "We'd be freaking out on Opportunity right now," Jim says, looking at me suspiciously.

Ray (participating remotely, via conference call) reassures Jim that we do this all the time. Jim still looks worried even after Ray's reassurance, but then, Jim always looks like that. As the CAM progresses, Jim turns to me and asks, sotto voce, if I've worked on Opportunity yet. I tell him about my one chance to do so, which was lost because they were in a restricted sol and didn't need me. "You should work on Opportunity," he says. He vows to talk to Arthur about it when Arthur returns from travel. Fine by me!

I should have paid more attention to the drive sequence and less to my email. Jake Matijevic catches something I should have caught: since yestersol's drive failed, the rover set its goal-error flag, and we neglected to clear the flag in thisol's drive sequence. Thankfully, it's a trivial change, but I'm ashamed of missing it. (Normally I check the downlink report to see if any flags need to be cleared, but because of the late downlink the report hadn't been written yet when I checked, and I forgot to recheck it later. But I should have thought of it anyway, downlink report or no downlink report. Sheesh, how long have I been doing this?) And I don't feel any better when Ray says, "Good thing you were there, Jake," in a tone that makes it clear that, shall we say, he agrees I should have caught it.

He's right, though, that it is a good thing Jake was there: we already blew our only contingency sol, and this rookie mistake would have cost us another. If he hadn't been on the alert, we'd have lost a significant pre-conjunction science opportunity.

I reflexively beat myself up over this, but after reflecting on it a while, I start to feel less bad. This is why we have teams, after all: you're going to fuck up once in a while, and other people are there to backstop you. Which makes me realize that I haven't been doing my part as a team member -- I tend to do other things during the parts of the sequence walkthrough and CAM that don't directly affect me. So I decide to pay more attention during these meetings, in the hope that I can pay Jake's favor back, or at least pay it forward.

When the CAM is over, I start looking ahead at the schedule. From now until solar conjunction, I'm the only rover driver scheduled for Spirit. There's a floater scheduled for each of those sols, but it's not the same as having a second RP who's exclusively dedicated to the same rover. And they're going to be unusually important and challenging sols, with precision driving, and likely IDD on each sol, to boot. We don't have any more contingency sols to spend, and if we get something wrong, we could actually lose the vehicle: if I manage to put Spirit in a poor solar attitude before we lose contact during conjunction, she could drain her batteries, fail to heat, and the cold could kill her.

It's all going to be up to me.

It's a good thing Jake is there.