Mark Adler

State from Sol 11:

• The DTE session was interrupted by wind problems, and the DSN antenna was stowed. Nevertheless, the most critical Hazcam images from the middle and the end of the traverse were received. That was sufficient to determine that the approach was successful, and rover ended up just in reach of some targets on Kaibab, though apparently not the most desirable ones. The spacecraft is in good health, and all functionality is available. This is a 34m Sol, and so the DTE is extremely limited. 4.05 Mbits, minus 2 Mbits for engineering. Not even enough left for one Hazcam pair. More on that later. The UHF pass was received, and parts of it were used during sequence development. More on that later as well.

Sol 12 is an IDD Sol:

• It was decided that the potentially accessible targets on Kaibab were adequate to not require another approach Sol. So Sol 12 is an IDD Sol. Through the SOWG meeting, the accessibility of the potential targets was being developed and was not finalized even by the end of the SOWG meeting. So the SOWG chair had to go to the sequence meeting not knowing what targets we could get to with what, and so he carefully established priorities with the science team to be able to make target decisions during the sequence development process.
• The key activity is to get at least two of CMI, APXS, and Mössbauer on a target in reach, and furthermore to get enough status in the DTE to determine that at least two of the measurements were be likely to have been successful in order to permit a long traverse on Sol 13.

Sequence Development:

• The sequence development for this Sol was, well, interesting. Coming into the sequence meeting, we had only one target on which we could do APXS and Mössbauer, but not CMI. This would make it tricky to get confirmation in the DTE (but perhaps possible if only X-ray were done with APXS, as was pointed out by Steve later). Despite the failure up to that point to find a CMI target, the mission manager sent off one sequence developer (Chris) to find an accessible CMI target and to design that subsequence, in parallel with the main sequence development. The main development would plan the two instrument in situ observation, but we would insert the third (CMI) if available during the development process.
• A large sequence was developed to take sky mini-TES observations, cloud movies with the Navcam, three different series of IPS measurements (some that were repeats from a previous Sol), Pancam mosaics of Kaibab and Bonneville, and the arm deploy with APXS and Mössbauer measurements. Near the beginning of the sequence development, imagery that was supposed to be on the DTE was received on UHF that was intended to support the observation development. The mission manager approved the use of that imagery to adjust the placement of the Sol 12 observations. Additional Pancam on Kaibab and Bonneville was added to fill the FIDO time available, as well as make better use of the UHF.
• Chris succeeded in developing a CMI observation after some contortions of the arm to avoid hitting the rock. The CMI was on a spot not coincident but relatively near to where the APXS and Mössbauer measurements were being taken. The mission manager then sent Ashitey off to verify the safety of the arm movements designed by Chris. They came back with a verified plan, and the CMI observation was integrated with the other arm operations. This sequence was incorporated, but the ordering of it and the spectroscopy did not permit getting the status on DTE. So Chris was sent back to reorder the observations: CMI, Mössbauer, and APXS. He came back with that in time to integrate it into the overall sequence before the validation phase of the sequence development (see the modified sequence development schedule in the comments below). It was integrated successfully. Chris Leger was the hero of this Sol.
• Now we were left with the problem of getting enough information on the DTE to permit a drive on Sol 13. That information is both Hazcam pairs from the CMI placement and the Mössbauer placement (two out of three in situ observations are required for a location). Unfortunately the DTE volume would not permit even on Hazcam pair, though just barely. The test director was asked for a ruling to allow more highly compressed Hazcam pairs to permit fitting both pairs in the limited DTE, which meant more than a factor of two higher compression from the standard 16:1. It was asserted and supported that such high compression levels with the ICER algorithm would permit determining what would be needed to ascertain the efficacy of the arm operations. The test director permitted the compression approach (even though the images we will see won't really be compressed), and the sequence placed both Hazcam pairs on the DTE.
• During the sequence approval phase, it was noted that during the cloud movie, relative mast movements were being used for the first time. After some discussion it was realized that the last of a series of relative moves took the mast to exactly its limit at 270°. While absolute moves to 270° had been done before, relative moves had not and there was some concern that accumulated error might bust the limit. The consequence would be a termination of the sequence and no arm activity, since the arm movements followed the movie. The mission manager requested either a rapid redesign of the sequence to replace the final movements to 270° with absolute commands, or if not possible in the time available, to simply delete the movie sequence. The sequence redesign was accomplished, and that portion of the sequence was revalidated. The approval meeting then continued.
• The arm movement near the rock generated considerable discussion, but the in the end, the mission manager concluded that from previous FIDO testing and arm software testing, and from the quality of the range data that was used to plan the arm movements, that the operation was safe. We will see if the mission manger was right.
• The uplink was approved four minutes before radiation. Fun was had by all.

Sol 13 Considerations:

• Verify that the two Hazcam image pairs of the CMI and Mössbauer placements shows proper placement. Based on that (or the lack of it) make the judgment call that we can or cannot leave Kaibab. Verify that the Navcams from Sol 11 were received on UHF during Sol 12 to permit the drive design to Bonneville or to wherever the SOWG decides to head. Make sure everything is stowed before heading out, of course.
• We have 40m of traverse and one location so far. If the above two measurements are successful and the data are received (first attempt during the UHF pass during Sol 13 planning), then we will have two locations. Trenching is planned for Bonneville, if we go there and if it turns out to be what they think it is. A really long drive should be planned after Bonneville if we expect to make the 200m bogie. The system is capable of up to 100m a Sol over relatively flat terrain. Traverses of 100m to 200m have been suggested by members of the science team, in order to explore areas that cannot be seen from the lander panorama. This sort of discussion should be encouraged.

Cumulative Comments:

• One science request was not correctly recorded during the Sol 11 SOWG, and we instead implemented different version of the observation. This resulted in too many resources being applied to that observation that could have been used to do an atmospheric observation that had to be dropped. The SOWG chair must record and read back the decided upon observations for confirmation.
• We should note where we actually end up on the approach to adjust expectation for future approaches. I wonder whether halfway is the right choice, since there is probably a bias one way or the other based on the terrain characteristics.
• We have only gone a little over 40m so far. 200m mission success will not be met without very aggressive driving after Bonneville. We will need to define a long traverse over terrain so flat and obstacle free that we can safely turn off hazard avoidance. (This is probably a note for me on Sols 15-18.)
• We should change the FIDO timeline to document how this seems to be really working:

Is:

• 15 minute Sequencing Team Meeting
• 60 minute Build and Validate Sequence
• 15 minute Review and Approve Sequence

Should be:

• 70 minute Sequence Design
• 10 minute Sequence Validation
• 10 minute Sequence Approval

In addition, another 30 minutes of meetings follow at least the morning uplink at 10:30 am:

• SOWG debrief to summarize what they did and didn't get in the sequence development
• Selection of PIO image or other releases, and assignment to science team for creation and story writing
• Squyres debrief on science team performance and recommendations

• Three (!) targets were all named "arm8" by different people creating them. One was later renamed to "arm10\" (the "\" is not my typo). As it turns out, one of the remaining "arm8" targets was used in the uplinked sequence. Great care was taken to make sure it was the right arm8, but this should clearly be avoided if possible.
• When reusing a sequence from a previous Sol, ask if a) the state before and after the inserted sequence is set properly, b) were there any errors realized in the old sequence that need to be fixed, and c) were there any other assumptions about that sequence that need to be rechecked in the new environment.
• The mission manager forgot to note where the approach on Sol 11 ended up relative to the midpoint target, as he had recommended. Someone should.
• It seems that this mission manager likes to refer to himself in the third person.

Back to top