Mark Adler

State from Sol 15

• Rover in good health, all functionality is available.
• The Sol 15 sequence was terminated at 90 minutes with about 10 minutes of predicted sequence time to go. As a result two of the IPS observations and the second Hazcam pair was lost. The reason was that the FIDO time modeling of a 1000 coadd IPS point is correct, but the model is linear and apparently underestimates the actual time for a 2000 coadd IPS point. There were a lot of IPS points, all of which were 2000 coadds (except for the cals).
• There is some confusion on how well the rover needs to know its absolute heading. This matters most for images taken relative to true North. The Sun sensor has not been able to see the Sun for three Sols. (This is a FIDO artifact due to the limited field of view of the Sun sensor.) It has been reported by the atmosphere theme group that the absolute heading has drifted on the order of 20° to 50°. However on MER this would mean that we could not point the HGA at Earth, and therefore would have received no DTE! This is clearly not the case. There is therefore a contradiction in the evidence. For now the test director has said that we can magically receive DTE even if the rover doesn't know it's heading. To support image planning, we will try to image the Sun this Sol to reset the heading. That may be sufficient for image planning for a few Sols. For the next Sol, we will need a ruling from the test director as to whether the mission manager and spacecraft team need to continue to maintain heading with Sun observations in order to continue to get DTE. Note that the absolute heading is in general not needed for terrain remote sensing, traverses, or IDD operations, since those are done relative to images taken from the rover, and so are in a rover frame of reference.

Sol 16 is a Long Drive Sol:

• Six mini-TES observations requested 12 minutes into the SOWG meeting: Dessert Tortise, atmosphere 30° elevation East, Calico, Goose Prairie, Wenatchee, APXS Site 9. The last observation was later dropped, due to a decision to not do a small traverse to enable it.
• It was concluded that we could deliberately plan a drive over the 90 minutes of FIDO time in order to maximize the traverse, and depend on the drive contingency sequence to provide a 180° Navcam and front and rear Hazcam images at the end. The contingency is activated in enough time before the 90 minutes are up in order to complete. However in the end it was decided to not do that in order to get other important observations at the end of the drive.
• We will plan to take Sun observations before and after the drive to recalibrate the rover heading for the interpretation of images taken at the current site and to support planning for absolute observations after the drive.

Sequence Development:

• This is first and foremost a drive Sol to get as far as we can towards the target, which may be over 70 meters away, and to reliably end the sequence before 90 minutes to complete the activities at the end. We must determine the Sun direction before and after the drive.
• This development was dominated by the design of the traverse. Navcam imagery was used to plan the traverse. Pancam was available in the UHF, but was not deemed sufficiently registered to the Navcam data to use for traverse planning. This is the longest traverse yet planned in MER-FIDO tests. The total is approximately 55 meters, 35m to 45m of which is in terrain for which we do not have complete Navcam range maps. There is a significant probability that the rover will encounter obstacles in the latter portion of the traverse, and it will be up to the rover to either drive around the obstacles or to error out and run the contingency. So based on previous favorable experience, we are depending on the rover to take care of itself. Based on three independent estimates, the activities including the drive will complete in less than 90 minutes. The mission manager's estimate included an uncertainty, and showed a margin of five to 25 minutes (his estimate was 35 to 55 minutes for the entire drive). The mission manager concluded that the risk of prematurely terminating the sequence was acceptable, considering the margin and considering that the long drive contingency will provide the necessary information to continue the drive, and the Sun find can be redone at the beginning of the next Sol from the same position. This was the best approach to maximize the drive distance while still providing a good probability of performing the Sun find at the end of the drive.
• The mission manager named one of the traverse waypoints "adlers folly".
• The uplink was approved with 2.5 minutes to spare.

Sol 17 Considerations:

• How far did it get? Where did it end up? If the drive was not prematurely terminated, how long did it take? How does that compare to our estimates? If it did terminate prematurely, why? Can we still see our target? Do we have good bearings on the Sun direction before and after the drive? Do we need to redo the Sun-find at the beginning of Sol 17? Will the test director require absolute bearing knowledge in order to send DTE?

Cumulative Comments:

• It is important for the theme groups to sequence their requested observations, and in particular they should sequence their highest priority observations first. On Sol 15, the highest priority Pancam was not sequenced by the theme group, but the second and third were.
• Given extra time during sequence development, it is worthwhile to review the science priorities as well as to review, in sequence order, the visualizations of the image and remote spectroscopy requests.
• The modeling for IPS points needs to be updated. Ground testing should be performed to determine the duration of IPS measurements as a function of coadds. In addition, the modeling for Navcam images needs to be updated, since they appear to take up twice as much space in the resource modeling as they should at 16:1 image compression after a 12:8 pixel compression.
• We need clarification from the test directors on the requirements to receive DTE, i.e. whether the spacecraft team needs to maintain absolute heading information, and to what accuracy.
• The tools do not provide a means to determine the actual length of planned traverse segments. Furthermore, there are no good estimates with uncertainty of the meters per minute that the rover (with hazard avoidance on) can traverse as a function of the terrain type. Finally, there is limited information on the dead reckoning capability of the rover on gyros with respect to heading errors. Overall, there was no way to estimate the traverse duration or error based on the traverse design and terrain expectations, other than the experience of spacecraft team eyeballing the situation. This would not be acceptable in a real mission.
• We are currently at 2 out of 3 locations, zero trenches, and approximately 60 meters. If all goes well, we should be over 100 meters at the end of Sol 16. At halfway through the 20 sols, that wouldn't be too bad.

Special Notes:

• Starting on Sol 16 and on all subsequent sols, the afternoon sol will begin at 12:15 PM with all subsequent events shifting later by 15 minutes.

Updated sequence development schedule starting at 0900 and 1415:

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

Back to top