2010-10-28

Opportunity Sol 626 (Spirit Sol 647)

Our autonav drive actually went great -- a total of 45 meters, quite respectable by current standards. That means we got about 15m out of the autonav, which is good news.

The only fly in the ointment is that we failed to do our turn for comm, and for a reason we should have thought of. When we planned this drive, we planned for it to extend beyond the time-of-day limit. This is common -- indeed, something we used to do all the time when using autonav; we just have it drive until the time-of-day limit kicks in, then we have the sequence detect that we were stopped by the time-of-day limit, clear the error, and reset the limit to allow time for end-of-drive activities such as turning to comm. We did all that logic on this drive, so we thought we were in the clear.

But here's what we forgot: both the blind and autonav portions of the drive are punctuated by slip checks, helper subsequences that try to drive a short distance and preclude further driving if we didn't make a decent fraction of the commanded distance. After the time-of-day limit hit, the sequence dutifully tried a slip check, noticed that it hadn't actually gone anywhere, and precluded driving. This is something our end-of-sol error-clearing logic didn't check for (indeed, has no good way to check for). So we still had an error -- the rover thought it was stuck in another dune -- and we didn't turn for comm.

Oops. We're still happy, but oops.

Well, our downlink was still pretty good. We didn't get the PCAMs, so that limits our blind driving thisol.

Hey, no problem. Jeng and I plan a sequence that drives blind as far as possible, then uses autonav until stopped by the time-of-day limit.

And this time, we do it correctly.

No comments: