The MER Science Process Steve Squyres

1
The MER Science ProcessSteve Squyres
2
The MER Science Process

• Introduction
• Operations Flowchart and Sol Trees
• Science Operations Roles and Responsibilities
• Some Rover Best Practices

3
Introduction

• The MER rovers are robotic field geologists.
• Field geology is an iterative process of
  scientific hypothesis formulation and testing,
  performed in a field setting. Most of us are used
  to doing it on foot, using simple tools and
  working alone or with a partner.
• For MER, we have had to develop new processes for
  dealing with the unique challenges of doing field
  geology on a distant planet with a robotic
  vehicle
• Learn how best to live with the constraints and
  weaknesses of the system
• Limited mobility
• Limited data bandwidth
• Limited soil/rock manipulation capability
• Navigation errors
• Long response time
• Learn how best to take advantages of the unique
  strengths of the system
• VIS/NIR multispectral imaging
• Remote thermal IR spectroscopy
• In-situ APX and Mössbauer spectroscopy
• Many experienced scientists working together as a
  team
• And while all this is going on, we also have to
  do outstanding atmospheric science.

4
Introduction (contd)

• Initially (3-4 years ago) we knew very little
  about how to do field science with a rover that
  has MER-like capabilities.
• Our key learning tool has been the FIDO rover.
  This vehicle is MER-like in many respects, and
  has allowed us to experiment with how to do
  robotic field science.
• In a series of three field deployments, we have
  developed and refined our science operations
  approach and processes
• 1999 First field test. Identified basic sol
  types, and formulated first operations flowchart
  for an in-situ science/sample collection mission.
• 2000 First blind field test. Developed basic
  science operations processes and refined the
  operations flowchart for a pure in-situ science
  mission.
• 2001 Second blind field test. Greatly refined
  science operations processes, refined operations
  flowchart (addition of touch-and-go sols),
  developed first sol trees.
• We have also collected a list of rover best
  practices throughout this process.

5
Operations Flowchart and Sol Trees
6
(Navcam pan)

• Panorama Sol
• 360 color stereo Pancam
• 360 Mini-TES
• MI
• Nav/Haz (if necessary)
• Mössbauer (day)
• APXS (night)

New panorama needed?

• Drive Sol
• Spot PC / Spot MT / MI
• Touch-and-go APXS or MB
• Drive (Nav/Haz support)
• End-of-traverse Pancam
• End-of-traverse Mini-TES
• Nav/Haz

yes
no
Science target w/in 10 m?
no
no
(full PC/MT pan)
yes
Next science target spotted?
yes

• Approach Sol
• Spot PC / Spot MT / MI
• Touch and go APXS or MB
• Approach target
• Nav/Haz
• Pancam images of target
• Mini-TES image of target

no
In position?
(full PC/MT pan)
(Hazcam)
yes
yes
(spot PC/MT, APXS/MB/MI if available)
Done with this target?
Prepare target?
no
no
(APXS/MB/MI, spot PC/MT)
yes

• Scratch and Sniff Sol
• Spot PC / Spot MT / MI
• Prepare Target / Front Hazcam
• Target MI
• Target Mössbauer (day)
• Target APXS (night)
• Hazcam support

• Spectroscopy Sol
• Spot PC / Spot MT / MI
• Target Mössbauer (day)
• Target APXS (night)
• Hazcam support

7
Notes on Operations Flowchart

• Most of the MER surface mission will be a 90-sol
  traversal of this flowchart.
• The activities listed for the five sol types form
  a template for each sol. Not every one will
  necessarily be executed (in fact, resource
  constraints or other circumstances often will
  prevent all of them from being executed).
• Decisions are made on the basis of the data
  obtained to date, considering the hypotheses
  being tested and using the combined scientific
  judgment of the team.
• Atmospheric science observations are not included
  in the flowchart, but will be inserted into each
  sol as appropriate. Any sol type can, in
  principle, support atmospheric observations.
• Some special-purpose activities also are not
  included, e.g.
• Calibration campaign and egress
• Soil excavation (e.g., trenching with wheels)
• Thermal inertia measurements

8
Sol Trees

• When implemented sequentially, the flowchart
  leads to a sol tree

Sol N
Panorama Sol
Sol N1
Approach Sol
Drive Sol
Sol N2
Drive Sol
Approach Sol
Approach Sol
Spectroscopy Sol
Sol N3

• The following slides show some actual sol trees
  from the last FIDO test. (Transcribed from
  whiteboards/flipcharts to electronic form by
  Michael Sims.)

9
Sols 5-9
(Distance, Target, END OF DAY Locations,
VALUES PanCam)
SS Aaron
Sol 5 Sol 6 Sol 7 Sol 8 Sol 9
Approach Guild
Approach Guild
MI,MO,APS Guild
(13, 2, 0, No)
Begin Long Drive 1
MI,MO,APS Guild
(50, 2, 1, No)
(13, 2, 0, No)
10
Sols 9-13
(Distance, Target, END OF DAY Locations,
VALUES PanCam)
(13, 2, 0, No)
Begin Long Drive 1
MI,MO,APS Guild
(50, 3, 0, No)
Sol 9 Sol 10 Sol 11 Sol 12 Sol 13
(63, 3, 0, No)
(50, 3, 1, No)
Begin Long Drive 1
Begin Long Drive 2
(130, 5, 3, No)
(63, 3, 0, No)
Begin Long Drive 3
2 Sol Pan
(170, 6, 4, No)
(63, 3, 0, No)
Begin Long Drive 4
2 Sol Pan
Low data
(63, 3, 0, No)
(120, 3, 0, No)
Begin Long Drive 3
(170, 6, 4, No)
Spec - get location
2 Sol Pan
11
Sols 13-17
(Distance, Target, END OF DAY Locations,
VALUES PanCam)
(63, 3, 0, No)
(63, 3, 1, No)
Sol 13 Sol 14 Sol 15 Sol 16 Sol 17
Begin Long Drive 3
(120, 5, 3, Yes)
Spec - get location
Begin Long Drive 3
(120, 5, 3, Yes)
(63, 3, 0, No)
Engineering - fault recovery
(63, 4, 1, Yes)
Retry Spec - get location
Begin Long Drive 3
(120, 3, 1, Yes)
(170, 3, 0, Yes)
Begin Long Drive 3
(120, 4, 1, Yes)
Retry Spec - get location
(170, 4, 1, Yes)
Begin Long Drive 4
12
Science Operations Roles and Responsibilities
13
Terminology Tactical vs. Strategic

• Tactical refers to work that is necessary to
  get a final set of commands up to a rover on each
  sol.
• Strategic refers to work that supports the
  tactical process, but not on a single-sol time
  line. Examples include science data analysis
  over several days or more, data archiving, press
  conference participation, etc.

14
The Tactical Activity Cycle
15
Science Theme Leads (STLs)

• One set of STGs per rover. Team members and
  collaborators assign themselves to Science Theme
  Groups according to their interests
• Geology
• Mineralogy/Geochemistry
• Atmospheric Science
• Rock/Soil Physical Properties
• Long-Range Strategic Planning
• STLs are selected by STG members (except for
  Long-Range Strategic Planning).
• Each STG has a designated Documentarian for each
  sol, selected by STL.
• STL presents STGs activity request at SOWG
  Meeting.

16
Science Theme Group Members

• STGs continuously maintain a set of hypotheses in
  their area of interest. These are documented and
  tracked by the STG Documentarian.
• On each sol, the STGs receive quick-look data
  products from Payload Downlink Leads, and perform
  science analysis
• Discuss the implications of recent and past
  results for existing hypotheses.
• Evaluate the validity of the existing hypotheses
  develop new hypotheses as necessary.
• Identify observation concepts that can be used to
  test both old and new hypotheses.
• Convert observation concepts to preliminary
  activity requests.
• Present new findings, new hypotheses, and
  observation concepts/preliminary activity
  requests at Science Assessment Meeting.
• Based on presentations received from other STGs
  at Science Assessment Meeting, revise activity
  requests as appropriate, and prepare for
  presentation by the STL at the SOWG Meeting.
• Continue to discuss/formulate hypotheses offline
  from the tactical process.

17
Long-Term Strategic Planning STG

• Generates and continuously maintains a strategic
  science plan that is consistent with Mission
  Success criteria, on two different timescales
• 90 sols. This is a full-mission strategic science
  plan. Not sol-by-sol, but a high-level sequential
  ordering of key activities that seeks to maximize
  science return. (Example Sample diversity in
  current area, then climb nearby hill for
  panorama, then traverse 250 m to nearby crater
  rim, then sample rim deposits, etc)
• 1 week. This is maintained as a
  constantly-evolving sol tree. As each sol is
  executed, old branches are pruned and new
  branches are added, consistent with Mission
  Success and the 90-sol strategic science plan.
• Tracks accomplishments vs. Mission Success
  criteria. Checks the evolving 90-sol plan and
  each 1-week tree branch for consistency with
  Mission Success requirements.
• STG Documentarian has the lead responsibility for
  maintaining a comprehensive science log for this
  site. Merges hypotheses from all other STGs, and
  continuously tracks hypothesis status, approaches
  to hypothesis testing, key scientific findings,
  and generation of new hypotheses.
• STL helps represent science on the Mission
  Planning Team.
• STL coordinates science discussions that take
  place offline from the tactical process.

18
SOWG Chair

• One per rover per sol.
• Six people trained for this job. One backup
  always available at JPL for each rover.
• Leads Science Assessment Meeting. This is a
  loosely structured meeting where recent science
  results are presented and discussed.
• Leads SOWG Meeting, guiding SOWG to consensus on
  rover and payload activity plan for upcoming sol.
  This is a highly structured meeting, run on a
  rigorous schedule.
• Adjudicates in situations where consensus cannot
  be reached.
• Participates in Uplink Process meetings, helping
  to assure that the final sequence is as
  compatible as possible with the intent of the
  SOWG.
• Each SOWG has a designated Documentarian for each
  sol, selected by SOWG Chair. Documents the
  tactical process.

19
SOWG Meeting Structure

• Roll call Make sure all key players are there.
• Updates on rover and instrument state of health.
• Update on resource predicts.
• Update on status with respect to Mission Success
  objectives.
• Short briefings on activity requests description
  and rationale
• Engineering
• Each STG (Long-Term Strategic Planning STG also
  reports on progress with respect to strategic
  scientific goals.)
• Discuss and reach consensus on which type of sol
  (panorama, drive, approach, etc.) will be
  executed.
• Starting from the basic template for that sol
  type, fill in prioritized activity requests.
  Identify and justify all critical data products.
• Identify contingency and runout activities.
• End product is a Tactical Science Activity Plan.

20
Science Interactions

• Effective interaction among the various STGs is
  essential!
• All STG members (and especially STLs) should
  check the Long-Term Strategic Planning
  Documentarians science log daily.
• Off-site scientists who are about to return to
  JPL should also check the science log carefully
  before returning.
• During downlink assessment period (before Science
  Assessment Meeting), Long-Term Strategic Planning
  STL should visit each theme group, get a heads-up
  on what theyre finding, and report back to the
  Long-Term STG.
• All STLs should always be alert to findings that
  could be of interest to other STGs, and should
  communicate them promptly. Dont wait for the
  Science Assessment Meeting if you have something
  really important!
• After arriving on shift (slightly before Science
  Assessment meeting), SOWG Chair should circulate
  among all STGs
• Look for areas where communication is necessary,
  and facilitate it.
• Get a head start on the SOWG meeting by learning
  what the STGs have in mind, and by starting to
  formulate the outline of an activity plan for the
  upcoming sol.
• An informal science get-together after the SOWG
  meeting (involving all STGs and led by Long-Term
  Strategic Planning STL) can be the most valuable
  interaction of all!

21
Some Rover Best Practices

• (Lessons from the FIDO School of Hard Knocks)

22
Remote Sensing Science

• When you have image data from your current
  location, one of the most effective ways to use
  Mini-TES is to start the day with a fairly small
  number of individual spots or short linear scans
  across carefully chosen targets. Choose the
  targets on the basis of geologic relationships
  and choose them to test specific hypotheses. This
  can provide almost as much scientific information
  as a big Mini-TES panorama does, and it takes
  much less time. Similar things can be done with
  high-resolution Pancam spots.
• When doing small Mini-TES spots a target with a
  small angular size, only try it if you have
  Navcam images of the target from your current
  location. Otherwise, youre likely to miss, and
  waste the observation.
• In general, Mini-TES should be used in 20-mrad
  mode during panorama days. 8-mrad mode requires
  excessively long integration times for rapid
  surveys.
• When selecting targets for spot Pancam and spot
  Mini-TES observations, always specify a priority
  ordering. This information is necessary for the
  Integrated Sequence Team.

23
In-Situ Science

• On Approach Sols, it is often advantageous to do
  a touch and go. This is analogous to MI
  target-of-opportunity science, but is performed
  with the APXS or Mössbauer. Even with very short
  integration times (1 hour or less) the Mössbauer
  or the APXS x-ray mode can provide useful
  information on a target of opportunity.
• When scheduling nighttime operation of the
  in-situ instruments (APXS, Mössbauer), generally
  put Mössbauer before APXS. This allows the APXS
  to benefit from the coldest overnight
  temperatures.
• When overnight time and/or energy is limited,
  consider performing a short APXS measurement that
  is only long enough to provide adequate SNR in
  the x-ray mode. This will yield good numbers for
  most elements, and will take much less time.
• When acquiring MI images, use a focus slew
  whenever possible, not a single frame.
• When taking in-situ data on any high priority
  target, take the time to make sure that you have
  acquired the data successfully before moving on
  to the next target.

24
In-Situ Science, contd

• When resources allow, consider doing
  target-of-opportunity MI imaging on Spectroscopy
  Sols, on targets other than the main spectroscopy
  target.
• Always get remote sensing data (Pancam and
  Mini-TES) on any target that you have done
  in-situ science on. If this means that you have
  to do a short backwards drive first to get it
  within view of the remote sensing instruments,
  then do it. And if it's a target that you used
  all of the in-situ instruments on, don't leave
  the scene until you're sufficiently sure that you
  got good data.
• For any in-situ observations, always pick a
  primary and a secondary target if you can. This
  allows flexibility in the Integrated Sequence
  Team meeting.
• When doing both remote sensing and IDD work at
  the start of a sol, put the IDD work second,
  since it's more likely to error out than the
  remote sensing.

25
Support Imaging

• Whenever you don't know exactly where you are or
  exactly what you're doing, take the biggest
  panorama you can. It will probably save you time
  in the long run.
• Extensive support imaging is very helpful for
  determination of rover position and orientation.
  Whenever possible, a complete set of Navcam (360
  panorama) and Hazcam (front and rear) images
  should be acquired at the end of every rover
  move.
• A Navcam panorama should be acquired every 20
  TBR meters during a traverse. Whenever
  possible, each panorama should be 360 in extent.
• Hazcam and Navcam images should be acquired at
  the end of every approach day. Navcam images
  should consist of as much of a 360 panorama as
  possible, to verify the rover's position with
  respect to the target. Hazcam images also aid in
  determination of rover position. In addition,
  they can be used to verify that the IDD
  deployment volume is clear in case the rover did
  not get to the desired position (which would
  afford the opportunity to perform a start-of-sol
  Microscopic Imager sequence on the subsequent
  sol).

26
Support Imaging, contd

• Acquire Hazcam images at the end and 1 meter
  before the end of every traverse. This means you
  will always have images of the workspace volume
  in case you decide you want to analyze a target
  that ends up right in front of you.
• Whenever an in-situ instrument has been placed on
  any target, acquire Hazcam, Navcam, and/or Pancam
  images to verify proper placement. For overnight
  arm instrument observations, these images can be
  acquired either at the end of the preceding sol,
  or at the beginning of the following sol. For
  start-of-sol arm instrument observations, these
  images should be acquired immediately after
  initial placement of the arm on the target.

27
Target Selection

• There are two rather different approaches one can
  take to target selection in a very complex scene.
  One is simply to target one particular rock and
  hope when you get there that the geometry is as
  good as it looked from a distance. The other is
  to target a rocky region... a place with a high
  concentration of rocks of roughly the right
  dimensions. In this case, no one rock may look
  optimal, but if you get there and it's bad, you
  may have other choices very close by.
• It can be dangerous to pick a target rock that is
  right on the edge of an image gap. There may be
  an unseen adjacent obstacle that will prevent a
  proper approach.

28
Planning Ahead

• For each command cycle, you must command not just
  the post-uplink portion of the coming sol, but
  also the pre-uplink portion of the sol after
  that. Because you won't necessarily have all the
  information that you'll want to plan the
  pre-uplink portion of sol N2 if the rover is
  going to move on sol N1, some special
  considerations apply to what you can do.
• Here are the basic rules
• If you are certain that sol N2 is going to be a
  Panorama Sol, go ahead and start it first thing
  in the morning. (Note that whatever
  target-of-opportunity in-situ science you want to
  do that day will have to start after the uplink
  is over.)
• If sol N2 follows a Panorama Sol, a Spectroscopy
  Sol, or a Scratch Sniff Sol (none of which
  involve moving the rover), do whatever you want
  you have the information you need to do it.
• If it's the day after an Approach Sol, do the
  best you can with Pancam and/or Mini-TES based on
  prior panoramas. Use whatever onboard attitude
  knowledge you have to do the pointing. When faced
  with choices, you should tend to favor more
  distant targets over closer ones, since you're
  more likely to hit them.
• If it's the day after a Drive Sol (and it's not a
  Panorama Sol), you're going to have rather
  limited knowledge of the scene around you. Make
  your use of Pancam and Mini-TES similar to what
  you would do on a Panorama Sol.

29
Odds and Ends

• When trenching, place the front wheel of choice
  where you want to trench. Then lock the other
  five wheels and drive the selected wheel in
  reverse. Then back away from the trench with a
  backwards drive. This prevents the rover from
  driving over the material excavated from the
  trench.
• Determination of rover position after a traverse,
  as well as analysis of events that occurred
  during a traverse, can be aided significantly by
  searching for and identifying rover tracks in an
  end-of-traverse panorama.
• At times, you may find that you have excess
  bits i.e., communications bandwidth available
  but no power to take more data. In this case,
  there are still good ways to use the bits
• Send down Pancam (or MI or Navcam) images at
  lower compression ratios, getting better-quality
  images.
• Send down Hazcam movies after a drive. These are
  great for outreach, and can also help
  significantly with rover performance evaluation
  and localization.