Thisol I'm RP-1 on Opportunity, and Paolo's soloing as RP-2. But it's a straightforward IDD sol, and he hasn't had many chances to practice IDDing, so he asks if he can carry the bulk of it and I accept.
This is convenient for me, since I'm putting together a training presentation. At Sharon's prodding, I'm doing something Saina asked for, oh, I've lost track of how many months ago: trying to educate the rest of the uplink team about what rover drivers do, what all the mobility commands mean, and so on. This presentation has turned into something of a monster -- about 75 slides -- and I'm not done with it yet. So being able to turn Paolo loose, even part of the time, and work on this presentation, turns out to be an enormous help.
Of course, when he needs me, I watch over Paolo's shoulder and guide him through the process. But he's doing pretty well already.
When we catch a break, he asks me wonderingly, "How did you guys train originally?"
I tell him about the thread tests, the limited training sessions we had, and so on. But it wasn't really much. "If a new candidate rover driver came along who knew only as much as I knew when we landed, I wouldn't let him anywhere near the rovers," I say truthfully. "But we made it through somehow."[1]
How, I'll never know. Mars must have been in a good mood.
[Next post: sol 660, November 10.]
Footnotes:
[1] Sadly, five years on, this has only become even more true: it's gotten harder and harder to drive the rovers, because the list of things you need to know has expanded so much. Also, our training opportunities are more limited right now: Spirit's not driving, and Opportunity ops tends to be very routine, with limited IDD work and drives that all look much the same as we cruise across the plains. So it's become harder and harder to train new people, no matter how sharp and talented they might be (as both of our current trainees are).
One of the things Ron Avitzur talked about in the Graphing Calculator story was end-user testing and how it ultimately improved the product. He described it as a frustrating but enlightening experience, seeing how user intuition interacted with a non-intuitive process.
ReplyDeleteMaybe the rovers would be easier to operate (and easier to train drivers for) if everyone at JPL wasn't so smart.
A lot of the problem stems not from the rovers themselves, which are actually highly operable -- unlike most spacecraft, they were designed that way -- but from what we've learned while exploring Mars itself. We've learned so much about driving in slopes, on different types of terrain, and so on, that it's just plain hard to capture and communicate all the stuff an experienced rover driver knows. Among other things, I've set up a Wiki where we try to store and organize all this information, but it'll never get perfect.
ReplyDeleteIncidentally, most spacecraft are not designed to be operable, and operations really suffers from that. Our job would be just about impossible if operability hadn't been designed into the rovers by very smart people.
That sounds like an extremely interesting wiki. That brings to mind a somewhat separate issue--it seems like project wikis tend to be project-internal, which makes sense in many ways, but I've done a lot of learning both during my BS and after by perusing project-internal documents from older missions. I wonder if our wikis will ever find their way to the coming generations of engineers.
ReplyDelete