2009-04-19

Spirit Sol 105

Chris tells me there's a bet on: after Opportunity's record-setting drive, some of the Opportunity scientists bet that Opportunity will reach 1km total drive distance before Spirit. Chris took Spirit's side.

"That's unwise," I say. But Chris has a good answer for that. Once Opportunity makes it to its next target, Endurance Crater, the scientists are going to want to sit tight for a long time. Endurance is only about 700m from Eagle Crater (the one Opportunity landed in), so they'll have less than 1km on the wheels when they get there. By contrast, Spirit's next destination is still about 2.5km distant. It's the tortoise and the hare.

Apparently, the scientists who proposed the bet thought better of it later, and changed their tune. They still want to bet -- but only if they can bet on Spirit.

We'll be getting about 40m more of that 1km thisol. Our last couple of drives brought us close enough to Missoula Crater that we can get there in one more drive. We'll be doing a touch-and-go, a few measurements on a single soil target followed by a drive up the crater slope.

Chris already has the IDD sequences built and is working on the drive. This is my first solo RP-2 shift, and I get right to it. Chris's sequences are sloppy -- he's got a nonzero APXS overdrive after a MB touch on a soil target, he's not minimizing IMU on-time by running the instrument start/stop sequences when the IMU is off, the MI Z-axis retraction move after the MI series is 77mm when it should be 80mm. In other words, it's about like what I used to hand to my RP-2s. I sigh deeply. "I'm not a detail person," Chris says.

Neither am I. But this is one of the joys of being an RP-2: you get to criticize the work of others, while doing less yourself. I'm used to being on the other side of that relationship, so this should be good for some petty revenge, at least. Until it drives me insane.

I don't have much to do except fix up Chris's sequences, work that continues after he leaves. I wrap up everything and deliver, then move on to working on the uplink report and movies. While I'm doing this, the Uplink Verification Lead, Carolina, says we're using the wrong sequence to start the MB integration. We check, and sure enough, she's right. So I have to redeliver. Naturally, we're using a new uplink process, and nobody's exactly sure how I'm supposed to do this. But we limp through it.

One of the fun parts of my day is showing the animations during meetings. As RP-2, I still get to do that. The RP-1 generally shows the animations at the activity plan approval meeting; the RP-2 shows them at the sequence walkthrough and again later at the command approval meeting. Thisol, since Chris separated the IDD and the drive sequences, I show the IDD sequences at the APAM and he shows the drive; I show both sequences at both of the other meetings. One of the reasons this is fun is that I get to play with camera angles and stuff, a little like being a movie director. For the IDD sequence, I choose a dramatic low-angle shot, looking up at the rover from a position near the soil target it's investigating. It can be hard to make straight-line drives interesting, so today I show the drive from the endpoint looking back toward the start. Sadly, I sometimes think my art is wasted on this audience.

Since Chris is gone, I can use both systems to display the two animations during the CAM. Wheeling my chair back and forth between the two machines reminds me of when I used to do something similar in the computer lab at my undergraduate institution. When nobody else was around, I'd set up fractint (a Mandelbrot Set explorer) on several machines, then roll from system to system, looking at the results of whichever systems happened to be finished while the other machines continued to chew on the math. This is a lot like that, only the machines are a thousand times faster now, and I get paid for it. And it's for real.

Also, the chairs are more comfortable.

As we're convening for the CAM, Squyres walks in, shaking his head. "What's wrong?" Mark Adler asks. "We're getting into the worst part of a planetary flight project," Squyres says, "the part where they write papers."

Adler finds another problem in the sequences during the CAM -- we set the mobility time-of-day limit too early, so if the drive timed out, we wouldn't recover the right way. We have to fix that as well, and right after I deliver the fix (my second redelivery of the day), Carolina asks another question.

"We were taking the autonav images every three meters before," she says. "Are we still doing that?"

Oops. After the FSW upload, one of the arguments to a command we send before every drive was supposed to change from 5 (which caused the old setup to take an image every 3m) to 1 (which is how we get the same result with the new software). I knew this and have been doing it, but Chris forgot -- and I didn't check it.

So I have to fix it and redeliver again. At this rate, I'm going to set another record, but it won't be one I'm proud of.

So we check the changed sequences. (Happily, I wasn't the only one who needed to change something.) Everything is now officially OK.

Then I realize I didn't check something Adler asked about earlier -- do we have any drive or IDD error flags set that we need to clear? That information is usually in the downlink report, but for some reason there was no downlink report thisol, and I didn't get around to checking the telemetry myself. With a sinking feeling, I walk over to the SMSA and check the telemetry. I do not need to go through the redelivery hassle again. Fortunately, the error flags are all clear: the vehicle's fine, and we don't need to change the sequences.

I continue working on the uplink report while everyone else goes home.

And, of course, I discover a problem. After a drive of this length, we really should increment the site index -- our way of telling the rover it's in a fundamentally new place. But somehow Chris left that command out of the sequence (or accidentally deleted it), and I didn't notice, nor did anybody notice during the reviews.

This isn't a serious problem, so I debate about whether to even do anything about it. There's no threat to the vehicle, but it will make life inconvenient on the ground, since we'll have to do more work to keep tomorrow's images separate from the current ones.

I don't think Adler will want to change the sequence again, but I decide he should at least know about the problem and get a chance to make the decision. So I call his cell phone (he left an hour ago) and tell him about it. Somewhat to my surprise, he decides we should make the change -- we don't have to uplink the sequences until tomorrow morning, so we can CAM and deliver the sequences then. I guess it's a good thing I called him, after all. So I also call Chris Leger to bring him up to speed, and build a changed version of the sequence so that they can just bundle it tomorrow.

I'm about to go home when Walter Goetz bursts in rather dramatically. For some reason, the Phobos imaging sequence was kicked off 15 minutes late, and they missed the moon entirely, wasting 30Mbits (a significant chunk of downlink) in the process. He's agitated and looking for someone to complain to. I know the feeling, so I try to lend a sympathetic ear. "It's not your fault, of course," he says, "but it's very frustrating, yes?"

Yes.

No comments: