Techical Update from Alex Foessel

1. Robotic rock classification

Last night, Nomad completed static acquisition of images and spectra of the
selected set of rocks and meteorites.  A highlight of successful performance:
Nomad classified rocks brought from Martin Hills (see side expedition info)
as terrestrial, despite the fact that they look very meteorite like.

Nomad drove from Camp Crickett to the Asteroid Belt, very close to Patriot
Hills, a place identified as a good meteorite search and classification
demonstration. Teleoperated distance driven: 1.5 km.

Nomad is running fine out the APU (auxiliary power unit) and the circuit
breaker on the generator tripped once when the robot was challenged to drive uphill
in deep snow. Resetting the breaker enabled normal operations.

Nomad drove at its usual speed of 30 cm/s. The team is preparing for activities
during the day to use the most of the good weather we are enjoying right now.

2. Antarctic science activities

Deployment of temperature loggers (3) to monitor a total of 7 probes in the blue
ice for the rest of the season to study the conditions that cause the rocks to
sink in the regelation portholes (refer to the pictures of this ice blue
features). This could explain the dearth of meteorites in the area.

3. Radar activity report

Acquired several (44) frames of radar data from a scene under absence of flying
snow to estimate the level of noise introduced by this phenomenon.

Took frames of a scene with elevations of the sensor from 0 deg to -14 deg every
2 deg on fresh snow to evaluate backscatter at different grazing angles.

Repeated last set of measurements to study backscatter on packed snow surface
(snow runway).

Simulated the displacement of a robot with a radar sensor moving it along 3 m in
steps of 25 cm towards a pile of snow of 50 cm high and a ditch 50 cm deep. The
depression angle of the sensor was 14 deg.

(For all of the above activities the radar is 1.6 m high).

Next activities: perform grazing angle test over blue ice, surface that might be
the most specular (on top of being spectacular!) surface.

4. INMARSAT station

Raised the antenna by 1 meter over the piling snow. The
signal improved dramatically and the connection is more reliable.

Report on Petrography Experiment by Pascal Lee


Human missions to Mars will have to include substantial capabilities for
doing science analysis in-situ, i.e. on Mars, since shipping many rock samples
back to Earth will be impractical. A key analytical capability will be to be able
to do petrographic analysis and to share petrographic information with
scientists back on Earth. By examining rocks in thin section, definitive
information can be derived about their nature, origin, history, and
possibly fossil content. The 1998-1999 field season of the NASA/CMU
Robotic Antarctic Search for Meteorites Program will be an opportunity
to conduct the first tests of a remote field petrology kit
developed by NASA Ames and the Mars Society.

Because of the dust generated and the need to protect samples from
contamination, rock-cutting on Mars will probably be an EVA
(extravehicular activity) or an activity conducted in an isolated environment
within a surface habitat. In the current Mars Society "Mars Arctic Research Station" (M.A.R.S.)
design, a glove case is envisioned in which rock cutting and thin section preparation
will be done through a glass bay. An astronaut would be inside the habitat's
shirt-sleeve environment and would manipulate  the samples to be analyzed
through pressurized gloves extending into the bay. The samples, if not
actually outdoors, may be kept in the bay under an environment close
to that prevailing outside.

The field petrography experiment package brought to Patriot Hills,
Antarctica, will allow the rocks studied with the Nomad rover's complement
of instruments to be characterized in detail while
the Nomad team is still in Antarctica. The kit will allow the team to cut
rocks in the field, polish them, make thin sections, look at these under a
petrographic microscope, capture interesting microscope views on video, and
share the data with collaborators back at NASA Ames and at the Mars Society
through the Internet. Rapid feedback and any needed science support would
then be available to the field team.

The field petrography kit comprises a "Gryphon" electric diamond band saw
cooled with propylene glycol (antifreeze) by capillary contact, a "Hi-Tech"
diamond grinder/polisher lubricated by propylene glycol, a Ward's petrographic
microscope, and a Mobile Science Workstation (MSWS) developed
by NASA Ames's Rick Alena of the Intelligent Mobile Technologies group
and his team. Rock-cutting and polishing will be done outdoors at Patriot
Hiils while the operator uses thick neoprene gloves to simulate the mobility
restrictions of spacesuit-type gloves.

The field petrography kit presently tested will be further developed during
upcoming field seasons of the Haughton-Mars Project at the Haughton impact crater,
Devon Island, Arctic Canada. While the current design of the field petrography kit is
human-operated, ways to automate the process in the context of a human
mission to Mars will be examined in consultation with the Carnegie Mellon University
Robotics Institute field team.

Support for this experiment was provided by
NASA Ames Research Center
The Mars Society
The Robotics Institute, Carnegie Mellon University
Fuerza Aerea de Chile (FACH), Chile
and is gratefully acknowledged.

Principal Investigator:
Pascal Lee, Ph.D., NASA Ames Research Center, MS 245-3, Moffett Field, CA

Co-Investigators/ Collaborators / Consultants:
Rick Alena, NASA Ames / Recom
Dimi Apostolopoulos, Ph.D., Carnegie Mellon Univ., Robotics Institute
Geoffrey Briggs,Ph.D., NASA Ames, Center for Mars Exploration
Theodore E. Bunch, Ph.D., NASA Ames
William Clancey, Ph.D., NASA Ames
Alex Foessel, Carnegie Mellon University, Robotics Institute
Brian Glass, Ph.D., NASA Ames
Steven J. Hoffman, Ph.D., NASA Johnson
Christopher P. McKay,Ph.D., NASA Ames
Kurt Micheels, M.A.R.S. Project Manager/Architect, Mars Society
Liam Pedersen, Carnegie Mellon University, Robotics Institute
Theodore Roush, Ph.D., NASA Ames
Robert Zubrin,Ph.D., Mars Society

See also:

Expedition Report from William Cassidy

30. Nov. 25, 1998, Wednesday: Patriot Hills Camp

The day seems a little colder than yesterday, is overcast to partly cloudy, and there is a light to medium wind. All in all, however, a more adequate day in which to carry out research programs. Liam was able to get his equipment fired up during the morning. There is an unfortunate problem with the pan-tilt camera: it should register in three channels -- red, blue, and green -- but it not registering in the blue. Liam thinks he can get around this, however, with later laboratory simulations, so the work proceeds.

During the afternoon, Liam, Sib, Mike, Matt, Pascal, and Fernando were out at Nomad and Liam managed to get the spectrum of one rock, of the 40 he would like to obtain. Well, the first time is always the most difficult. They were back at dinner time with the generator, which had been misbehaving again. Moncho and Peta suggest changing it with the carburetor of the old one. They have looked at it and think it would require about an hour's work to do this. However, when they looked at the new carburetor they found a big chunk of ice in an intake hose, so maybe this was the problem ad it has been cured. We will see tonight.

We received news today that the British House of Lords has voted not to allow Pinochet to return Chile. Ordinarily, one would not expect this to have any effect on research in Antarctica, but the Chilean Air Force Commander-in-Chief has now postponed his visit to the South Pole and the base personnel here have been ordered to be out of here by December 12. The scientific groups therefore must be out earlier, with a choice of December 2 and 4. We will try for December 2, so this puts a constraint on our activities and will require some tight planning and good weather if we are to finish out program here in the time remaining to us.

Pascal, Liam, Sib, and I went out after dinner. Sib and I occupied the tent at Camp Crickett and Pascal and Liam worked around Nomad, getting spectra of rocks. Sib raw his computer 'sweeper' which controls Nomad in teleoperation mode and I tried picking out rocks on the screen that we could go to and try to identify by their spectra. Liam operated the spectrometer sensor and Pascal identified the rock types so we could see how close the classifier had come to identifying the rock.

We learned a lot about the system. The images on the computer screen were too dim to be able to tell much about the rock. This may be a problem with the computer display, or it may be a problem with the automatic iris of the high-resolution camera that takes the image. It is well known that photography is difficult in Antarctica, in part because there can be so much contrast between light and dark in the same scene. We are looking at generally dark rocks against a light background of snow and ice. It may be that the light background is overwhelming the scene and the auto-iris is closing down in response, making it impossible to see details in the rock. Another problem that slowed things down considerably is that Sib cannot receive a real-time moving image of what lies in front of and around Nomad. Instead, he had to maneuver Nomad on the basis of still-photo images, examining one and then moving Nomad forward; then getting another image to see how close he was to the target rock. In some cases, the scene changed completely between images and he became lost; in other cases he overran the target rock and had to back up. These are important things to know; the knowledge can be gained only through experience, and we were getting the experience. This knowledge should promote design changes in future generations (i.e., Son of Nomad).

After about six rock identifications carried out in this manner we decided we had learned enough and changed to a faster mode of operation in which Liam and Pascal picked out rocks to identify and directed Nomad to their vicinity via hand-held radios between us and them (we were about 2 km away). We had started about 1:30 a.m., and by 7:30 a.m., when we broke for breakfast, we had identified 13 rocks, one of which was a meteorite planted by Pascal in Liam. In each case except one, the classifier was able to tell the difference between a terrestrial rock and a meteorite. In the single exception, the rock's surface was rather rusty and it was identified as a meteorite with better than 50% probability. Liam hopes to do 40 rocks and 20 meteorites, so the work will continue.