I see Tatiana[1] on my way in, and she apologizes for the SpaceOps travel not happening. But I tell her it's no problem, which it isn't, and she looks relieved to hear it.
Jeng commiserates with my non-misery, too. "How did they get it so fucked up?" he asks rhetorically. "Well, I'll tell you one thing, it takes pros to fuck things up this badly."
On my way up the stairs in 264, I see the TUL, Emily Eelkema, coming down. She cocks her head strangely and says, "You'll do."
It turns out that John went to the cafeteria 30 minutes ago and nobody's seen him. He doesn't have a cell phone, either -- somehow he manages to casually resist this. So they can't reach him, but they need to start planning.
In particular, they want to know if it's safe to deploy the IDD. I look at the imagery that's come down so far, and it's not promising. We didn't take the final stutter-step -- not too surprising, as there's a cluster of big rocks in our way -- and the HAFIQ[2], a set of images that's supposed to help us see what was under the rover in just such a case, turns out not to show us what we need to see thisol.
We've been in this fix before, and have been able to use thumbnail images[3] taken during the drive to show that there's nothing scary under the rover. So I go to the SMSA to see if Rich Petras and I can make the case.
Even without this problem, the rover's performance yestersol was a little disappointing. It made it through the blind drive just fine, but then the autonav got spooked by something. We can't see what spooked it, but we know we drove it up onto the top of a ridge, and our best guess is that the far side of the ridge sloped down too abruptly for a safe drive. Effectively, the far side of the ridge was an obstacle. So the rover started doing its little roverdance, making small turns as it steps back and forth, to try to find a way to the goal. In the process of doing this, it turned toward the big cluster of rocks we're currently facing -- another obstacle -- and, boxed in on two sides, never found a way out.
Mark Maimone tries to give me a hard time (jokingly) about the rover not completing its drive, but (equally jokingly) I'm having none of it. The author of the autonav code shouldn't have messed with me. "Oh, so you're saying we got through the blind drive -- that is, the part we planned -- just fine," I tease him, "and it was the autonav that failed. So this is our fault? Did we write the autonav code?" Mark mock-surrenders.
While Rich and I look through the thumbnails, John Wright shows up to do the same. We're pretty sure it's safe, but we're having a hard time working out the position of two possible deployment hazards -- we think they're well under the vehicle, or off to its right, but we're not sure yet. Art asks us for the status, and we tell him we're probably OK but we're still investigating.
"Bottom line, how are you going to convince me it's safe to do this?" he asks.
I shrug. "We'll have to convince ourselves first," I say. I think that was a good answer.
Ray Arvidson walks in to hear the tail of that conversation. "It's OK not to do this if there's any queasiness," he says. That puts a different face on the matter immediately. Rich, John, and I look at each other and shrug. "OK, let's not deploy, then," I say. Ray nods.
I should have expected something like this. Ray doesn't want to stop and IDD anyway; he's just going along with it because of demand from the other scientists. In the same vein, he enlists our help in selling this decision. "Don't be wishy-washy about it, though," he says. "Go into the meeting and say, 'We can't do this.'" That's John's job, and he's comfortable with it, so that's what we do.
The stress level drops rapidly, and we're able to joke about the situation a little more. Noting that the rover completed the blind drive and then autonav decided it wasn't safe to go any farther, Mark Adler says, "I guess the rover drivers knew exactly how far to go." Steve Collins starts doing a Kirk quote: "Risk ... is our business ...," he says, and I finish the quote: "That's why we're out there!" Adler laughs, but he says, "No, Steve, paranoia is our business."
Getting the rover out of here proves to be a small challenge. We can't drive over the rocks now in front of the rover. While they're technically under the rover's traversable limit, the combination of the rocks' own height and the size of the dropoff behind them effectively makes the rocks too tall. We can't drive into the zone that originally stopped the rover's traverse, because we can't see it in any of our post-drive imagery and all we know about it is that the rover thought it was a hazard. So we take the only remaining option: we back up. We can see that the area immediately behind us is safe, so we back into it, then drive over the ridge at a lower point off to our right before continuing. We have to ask MIPL for an additional terrain mesh before we can be sure this path is safe, but we plan the drive anyway, hoping for the best. If this doesn't work, we're out of good options. When the terrain mesh arrives, we use it to check out the drive -- and thankfully, it looks perfectly safe. So that's our plan.
As part of the ongoing cross-pollination between rover teams, Art is going to be Opportunity's Mission Manager during June. "Are you guys disappointed you won't have me to kick around?" he asks.
"No, but you're going to miss us," I reply. "You know what they say: absence makes the Art grow fonder."
Frankly, I'm lucky to be alive.
Footnotes:
[1] My group's secretary, who helps with travel arrangements and the like.
[2] The Hazard-Avoidance Fault Image Queue is the last several images taken by the hazard-avoidance (a.k.a. "autonav") code. If the rover decides to stop, it can be handy to see what led up to its decision; the images in this queue show you that.
[3] Scaled-down versions of the full images taken during the drive. These are small enough that we can often afford to downlink them at a relatively high priority; they don't squeeze out too much other data and can tell us a surprising amount.
No comments:
Post a Comment