 |
|
|
|
|
|
| Solar System Projects |
|
| Planetary Atmospheric Data Assimilation System (PI: Richardson, Mark,
California Institute of Technology
) | |
| We propose to develop a planetary atmospheric data assimilation system for the global extrapolation of spacecraft remote sensing
datasets. Planetary atmospheric datasets amenable to assimilation now include a multi-year dataset for Mars, and potentially mapping
data from Jupiter and Venus. Data assimilation improves the utility and interpretability of spacecraft data by essentially performing a
"four-dimensional retrieval": The physics inherently contained in a numerical model of an atmosphere allows an assimilation system
to globally extrapolate a sparse observational dataset and to convert observed fields (radiances) into physically consistent
meteorological fields (e.g., winds and temperature). However, it should be emphasized that the information content of the data is not
increased by data assimilation - the procedure simply maximizes the interpretability and information extraction from the data. As a
technique, data assimilation has been used for study of the terrestrial atmosphere for many years. The challenges for planetary
application of data assimilation include the very sparse observation patterns (compared to those of the Earth's atmosphere) and the
lack of an extensive, empirical climatological database with which to guide the relationship between the observed fields and model
variables. Consequently, "off the shelf" assimilation schemes cannot be directly employed in the planetary arena. Hence, we propose to
implement and test an ensemble-based scheme, which has the advantage of requiring minimal empirical knowledge of the target
atmosphere and better exploiting sparse observational sampling. Such techniques have shown great promise in terrestrial data
assimilation (though not widely or regularly employed operationally). Technical challenges include algorithm refinement and testing,
efficient implementation of the assimilation system, and various improvements in the forward numerical model. This forward model
will be the planetary version of the NCAR Weather Research and Forecasting model, previously developed with AISR funds. This
project will be a collaborative effort between Caltech and NCAR.
| |
| A Distributed Computing System Supporting Near Earth Asteroids Research (PI: Tricarico, Pasquale,
Planetary Science Institute
) | |
| We will develop a Distributed Computing system dedicated to Near Earth Asteroids (NEAs) research. This system will incorporate a
mature tool for Celestial Mechanics computations (ORSA) into a well established Distributed Computing infrastructure (BOINC), and
will be used in two different problems. The first application of this system is in the improvement of the performance of NEAs surveys,
by implementing a search strategy that maximizes the volume covered in the NEAs orbital elements space. This produces a very
accurate estimate of the unbiased NEAs population, and uses the entire set of astrometric observations of reference surveys to
determine the optimum scheduling that maximizes probability to detect new NEAs. The second application is the monitoring of the
impact hazard by NEAs, improving on standard systems in terms of response time, accuracy, and accounting for non-gravitational
effects. It uses a distributed version of the Monte Carlo sampling method, which tracks the dynamical evolution of virtual asteroids to
determine the statistical probability of an impact with Earth. This proposal addresses the AISR goal to demonstrate the degree of
relevance, applicability, and potential impact of distributed computing (an emerging information technology) to the NEA SMD
program. This proposal also addresses the NASA strategic sub-goal 3C: Advance scientific knowledge of the origin and history of the
solar system, the potential for life elsewhere, and the hazards and resources present as humans explore space, by advancing our
knowledge about the NEAs population and its impact threat.
| |
| Advanced Visualization in Solar System Exploration and Research (ADVISER): Optimizing Science Return From the Moon and Mars (PI: Head, James,
Brown University
) | |
| We propose to investigate novel information technologies and computational methods to increase productivity of NASA/SMD
research and exploration endeavors, and to extend the state-of-the-practice in those endeavors to help define SMD requirements for
return of humans to the Moon. This will be accomplished through use of ADVISER (Advanced Visualization in Solar System
Exploration and Research), an AISR-funded problem solving environment (PSE) for planetary geosciences that integrates and extends
state-of-the-art in hardware/software technologies and applies them to significant problems in science and exploration. ADVISER is a
set of tools providing planetary geoscientists with the capability to operate and analyze data as if they were on a planetary surface and
has four basic parts: 1) Geoscientist on the Surface: Visualization capabilities enabling placement onto the surface through immersive
virtual reality (IVR) and related desktop capabilities; 2) Importation and Visualization of Multiple Data Sets: On-demand importation,
co-registration and overlay of relevant image format data sets enhancing the ability to correlate/interpret data for scientific analysis; 3)
Field Kit Development: A menu with capabilities commonly carried out by field geologists (compass, altimeter, etc); 4) Ancillary
Virtual Field Instrument Development: Virtual items such as cameras, GPS, PDA equivalent of the field notebook. We propose three
thematic areas to develop and extend the ADVISER PSE: 1) Exploring/Optimizing the Software/Hardware Spectrum: (Dealing with
terabytes by incorporating huge new data sets; exploring the hardware middle ground between desktop and IVR). 2) Data Fusion:
Enabling Exploration of Space and Time Dimensions: (Exploring vertical dimensions, active geological processes; developing the
insolation "Time Machine"). 3) ADVISER on the Moon: Preparing for In Situ Human Exploration of the Moon and Mars:
(Optimizing ADVISER PSE, expanding field notebook capabilities with the Tablet PC; Optimizing scientific performance of the
geologist/Astronaut in the field through simulated Augmented Reality Displays). Results will contribute significantly to NASA/AISR
goals.
| |
| Automated Digital Terrain Model Generation from HiRISE and LROC Stereo Imagery (PI: Broxton, Michael,
Carnegie Mellon University West Campus
) | |
| The Mars Orbital Laser Altimeter (MOLA) has significantly advanced the study of the Martian surface by providing Geologists with
a highly accurate elevation map of the entire planet. In addition, it now serves as the standard for geodetic control and co-registration
of data sets on the martian globe. However, its limited resolution (463m/pixel at the equator) has rendered it insufficient for detailed
studies of specific sites; e.g. geologic stratification and deposition analysis, or in the case of mission planning, landing site selection. In
situations that demand highly resolute digital terrain models (DTMs), dense stereo reconstruction from orbital imagery has been the
favored technique. Extremely high resolution images from Mars Global Surveyor (MGS) and Mars Reconnaissance Orbiter (MRO)
have yielded terrain models of substantially higher resolution (10m/pixel or less) than MOLA. This will also be true when the Lunar
Reconnaissance Orbiter (LRO) mission flies in 2009: DTMs produced by the Lunar Reconnaissance Orbiter Camera (LROC) will
complement the coarse global map of lunar terrain available from the Lunar Orbital Laser Altimeter (LOLA) experiment. High
resolution stereo models allow scientists and mission planners to resolve important surface details with an unprecedented level of
detail, however building stereo-derived DTMs has historically been a time consuming process. Modern photogrammetric workstations
such as SOCET SET incorporate only a modest level of automation; considerable man-power is still required to create artifact-free,
well-controlled data products. The substantial number of man-hours and resources involved in creating a DTM from stereo imagery
has meant that only a limited number of these data products have reached the scientific community. The number of available stereo
pairs far exceeds the number that can be processed using existing techniques, and this gap widens as technology improves and each
new mission sends back more raw data than the last. To address this problem, the Intelligent Robotics Group (IRG) at NASA Ames has
developed an automated stereo processing software, the Ames Stereo Pipeline (ASP), that is capable of generating high quality DTMs
from satellite imagery using a fully automated process. The ASP is being used in existing collaborations to generate DTMs from Mars
Orbital Camera (MOC), High Resolution Stereo Camera (HRSC), and Apollo Panoramic Camera imagery. In this project, we propose
to expand upon the capabilities of the ASP and to transition from operating in a proof-of-concept capacity to mission and science
focused production. In particular, we plan to increase the utility of raw data from the current HiRISE and upcoming LROC missions
by providing scientists and mission planners with timely access to automatically generated 3D surface models from the latest orbital
imagery as it becomes available.
| |
| Planetary Spatial Analyst (PI: Keely, Leslie,
NASA Ames Research Center
) | |
| We propose the development of the Planetary Spatial Analyst, a software tool that provides 3D visualization and advanced spatial
analysis technologies for use with planetary mission data products. These technologies include methods for classification, spatial
statistics, comparison, and computation of derived data products as well as 3D visualization in the context of planetary terrain. With
the substantial increase in the volume of planetary data, traditional analysis methods have become time-consuming, inefficient, and not
easily reproducible. It is essential that researchers have advanced interactive and automated techniques for extracting information
from these data at their disposal. We will adapt many of the spatial analytical techniques currently used in the Earth science
community as well as some from the planetary science field. The Planetary Spatial Analyst will provide these techniques in planetary
coordinate systems and with software practices familiar to the planetary science community. Our tool will use the topographic data
products derived from orbital imagery for terrain visualization. Furthermore, this tool will be implemented as an extensible component
of the Mars mission software environment and will provide capabilities for collaboration using common geographic software
applications. We anticipate that the Planetary Spatial Analyst will accelerate fundamental knowledge building research in preparation
of the exploration of our solar system.
| |
| Multi-Modal Image Registration and Mapping for Titan Balloons (PI: Matthies, Larry,
Jet Propulsion Laboratory
) | |
| Over a three-year effort, the research proposed here will facilitate registration of Cassini synthetic aperture radar (SAR) and
visible/infrared (VIR) imagery and address needs of a future Titan balloon mission by developing and characterizing the performance
of algorithms for: - Autonomous registration of SAR and VIR imagery from flyby, orbital, and balloon platforms to improve science
image registration and to provide global balloon position estimates; - Autonomous topographic mapping, orthoimage generation, and
incremental balloon motion estimation with imagery from the balloon to improve image registration, data compression, and balloon
navigation. We will also show that these algorithms are amenable to onboard implementation with space-qualifiable computing
hardware that is already in development based on field-programmable gate arrays (FPGAs). Experience from Cassini shows the
scientific value of accurate registration of SAR and VIR imagery. Registration error of 0.2 to 0.5 pixels is desirable, but may not be
guaranteed by spacecraft position and attitude knowledge. Half pixel or better registration error has been demonstrated in limited
experiments on automatic registration of SAR and VIR imagery from Earth orbiting satellites, using algorithms developed for
matching multi-modal medical imagery. Balloon position knowledge is required for a number of reasons, including guidance to desired
sampling sites. Radiometric navigation can give accuracy around 100 meters (m) for a balloon, but with latency of many hours, which
limits its value for real-time guidance. Celestial navigation based on sensing directions to the sun and Saturn could operate in real-time,
but with error of multiple kilometers. Image registration has the potential to provide real-time balloon position knowledge with error
comparable to the image resolution; i.e., approximately 500 m for Cassini flyby data and 100 m for a future orbiter. Regions of Titan
seen to date generally have low relief, but mountains about 1 km high do exist. Creating topographic maps onboard the balloon and
creating orthoimage mosaics from the maps for registration to orbital imagery will improve registration and balloon position
estimation. Substantial steps have already been taken toward demonstrating feasibility of doing this onboard, in real-time, with
flight-qualifiable FPGAs. Doing these functions onboard will enable several other benefits, including: - Providing topographic maps for
science of the image swath seen by the balloon at resolutions ten times better than possible from orbit; - Reducing downlink data
volume requirements by roughly a factor of two by doing mapping onboard instead of on the ground; - Detecting terrain hazards as a
byproduct of real-time onboard mapping; - Detecting changes in real-time, for example to detect active cryovolcanism, as a byproduct
of real-time balloon position knowledge at meter scale derived from the mapping process. JPL is developing a high performance, low
power, space computing capability based on FPGAs as a coprocessor to standard rad-hard computing architectures. We have already
implemented some parts of these algorithms in real-time in such FPGAs, which gives us confidence that the entire set is amenable to
real-time, onboard implementation. Doing this research now will make the results available in time for a potential Titan Explorer
technology decision gate early in the next decade. This work can also benefit a Venus balloon mission and small bodies missions. Our
team has developed prototypes of the proposed capabilities and inserted onboard computer vision capabilities in previous missions,
particularly the Mars Exploration Rovers. The team includes leading participants in remote sensing of Titan and deep space navigation
to ensure relevance of the work and application of the results.
| |
| Detecting Transient Surface Features with Dynamic Landmarking (PI: Wagstaff, Kiri,
Jet Propulsion Laboratory
) | |
| Our goal is to dynamically and autonomously detect transient surface features, such as dust devil tracks or dark slope streaks on
Mars, from images. This goal is particularly timely given the announcement that recent surface changes (a new gully and 20 impact
craters) have been discovered on Mars (Malin et al., 2006). These critical discoveries tend to proceed slowly because scientists rely
primarily on manual examination of each image pair to identify any changes. The automated methods that do exist focus on per-pixel
changes, require full image registration, and are prone to generating false detections. We propose a major paradigm shift by
decoupling change detection from the individual image pixels. We will provide a landmark-based analysis that enables both easier
interpretation and the ability to quickly detect any changes. Dynamic landmarking can detect transient features without requiring
image registration, which represents a large step towards enabling the onboard use of this technology. Because the landmarks are
represented at an abstract level, it is possible to combine observations from different instruments at varying attitudes and under
different illumination conditions. No system currently exists to provide this level of sophistication in change detection for planetary
surface features; success in this area will greatly increase the number of surface chances that can be detected. We will demonstrate this
capability on overlapping images collected by the Mars Orbiter Camera (MOC) and Thermal Emission Imaging System (THEMIS),
two Mars imagers on different spacecraft. Dynamic landmarking first selects surface features that will function as landmarks in each
image. Landmarks may include craters, volcanoes, fissures, and so on, as well as the transient features we seek. Next, we extract
descriptive attributes for each landmark, such as size (surface area), shape, albedo, homogeneity, etc. Using these attributes, we can
automatically classify the discovered features into a set of known classes of interest--and mark any unclassified features as new,
potentially high-interest features. These landmarks provide a regional characterization of the area covered by the image that can be
used to better understand surface processes as well as to recognize when two images overlap. Finally, we can compare the landmark
sets discovered in two images to quickly detect any changes with high precision. One of the biggest benefits of this effort is the increased
productivity it will lend to science investigations, as compared with manual change detection. This work will lead directly to increased
science return from data that is already being collected. Specifically, it can increase our knowledge about other planetary surfaces, and
the dynamic processes present, to support future human and robotic exploration.
| |
| An Analytical Tool for Robot Mission Reliability Prediction (PI: Dolan, John,
Carnegie Mellon University
) | |
| Space missions require robots that are highly reliable for the smallest possible cost. Current theory and practice do not provide a
methodical approach to choosing a set of robot modules with demonstrably high reliability at minimal cost. In particular, there is no
methodology clearly linking mission specifications, to include desired reliability, to an appropriate choice of robot modules for
composing a robot or multi-robot team. The objective of the proposed work is therefore to establish a principled analytical
methodology for the inclusion of reliability considerations in mobile robot and multiple mobile robot team mission design. We will
additionally implement this methodology in software and make it available to the NASA and mobile robot community. This work is a
component of the longer-term project of creating comprehensive, principled analysis tools for general mobile robot design, to include
mobility, weight, speed, kinematic, and other characteristics. The proposed methodology involves three basic tasks: 1) define and
characterize fundamental mobile robot modular components; 2) develop a taxonomy for single and multiple mobile robots; and 3)
develop analytical and stochastic sampling-based methods for determining mission reliability. The proposed work addresses the
following particular goals of the AISR program. By quantifying the cost-reliability relationship in planetary rovers which perform
science exploration missions, it has the potential to increase the ¿efficiency of research endeavors conducted by Science Mission
Directorate research programs¿ and ¿mission duration and reliability¿ while ¿reduc[ing] mission development time, risk, and
cost¿through advanced simulation and design capabilities¿. The resulting system will allow NASA mission designers to rapidly
determine cost-effective and reliable modular robot configurations for a given mission and will serve as an aid in the cost-effective and
reliable design of the SMD¿s robotic exploration missions envisioned in the 2006 NASA Strategic Plan.
| |
| Tactical Rover-based Martian Geologic Mapping (PI: Crumpler, Larry,
New Mexico Museum of Natural History and Science
) | |
| Current state-of-practice techniques for location refinement of planetary surface rovers are too time-consuming to conduct during the
tactical science process of surface operations. As a result, uncertainty in the location of a rover during the tactical decision-making
process detracts from the quality of science planning. By expanding existing tactical software to "harvest" the inputs of team member
targeting of surrounding features during the normal course of science analysis, we have learned that it is possible to perform location
refinement quickly enough to inform the tactical science process. This improved knowledge will lead directly to the planning and
execution of superior geologic mapping observations. Enabling rover operators to perform location refinement during the tactical
planning process will enhance productivity in many ways, including: -increased data return and time available for execution of science
by using previously transmitted images for targeting and avoiding wasting time and bandwidth by retransmitting images of terrain
already on the ground to compensate for inadequate localization knowledge. -increased accuracy and quality of targeted science
observations -improved collaboration between science investigators, particularly those who are geographically distributed: We will
present an accurate geologic context to each scientist from within the science planning tool that he or she uses to analyze science data
and to collaborate with the group in planning new science observations. -support for integrating geologic mapping information from
rover science on a planetary scale by geographical information systems (GIS) integration¿opening additional avenues for cross-domain
collaboration and sharing of knowledge between science investigators This study directly addresses the specific goals of the ASIR
program by (1) adapting existing tactical processes to solve an important issue (positioning and data location within that position) for
rover science operations, (2) demonstrating that, through use of relatively simple improvements in existing data acquisition
methodologies the pace of mission accomplishments can be increased, (3) build on the interdisciplinary interaction of spacecraft
information specialists and end-users (scientists) to further refine operational methods, and (4) fostering increased science return
through methods that \"stitch\" site to site remote observations in a geologic framework supported by accurate position data.
| |
| Automated Orbital Mapping: Statistical Data Mining of Orbital Imagery to Analyze Terrain, Summarize its Characteristics and Draft Geologic Map (PI: Wettergreen, David,
Carnegie Mellon University
) | |
| Planetary exploration has entered a new era in which our data-gathering capability has outpaced our capacity for timely analysis. A
combination of high-resolution instruments, spacecraft autonomy and mobility, and extended lifetimes is adding to a vast library of
image and spectral data. The Mars Reconnaissance Orbiter is the latest following this trend, with HiRISE and MARCI images
providing unprecedented resolution of surface features. The catalog of images will continue to grow. The amount of resources devoted
to analysis remains mostly static, so we require automated methods to improve image analysts' productivity. We address the analysis
bottleneck in data analysis by creating methods to automatically generate maps from image and spectral data. Our method exploits
recent advances in machine learning and object recognition. The system will learn to infer mapping rules statistically from "training"
data in the form of existing geologic maps and registered orbital images. This is a natural application of existing learning-based
methods of image segmentation and boundary detection. We will extend these methods to the orbital mapping domain with
geologically-relevant descriptors like texture, spectral signatures, and surface features like boulders, craters, and dunes. The result is a
system that uses features to predict the locations of geologic units and their boundaries. Automating geologic mapping will yield
immediate benefits for the science community. It will dramatically improve the speed of image analysis which is limited by manual
feature labeling. Maps offer fast summaries of gigapixel images that may include hundreds of thousands of features. These "draft"
maps can be refined by human experts in less time than it would take to construct an entire map from scratch. Our mapping system
will also permit new classes of spatial data mining. Exhaustive statistical analysis of image libraries can quantify trends and discover
hidden correlations and anomalies. The system will yield summary statistics that improve geologists' ability to understand large
datasets rapidly. Automated mapping can also improve the quality of data analysis: it can quickly examine thousands of images,
perfectly, consistently and without fatigue. This reduces the need for redundant analysis for quality assurance. Automated mapmaking
is not intended to replace human expertise; both human and automated image analysis play to different strengths. The human can
offer expert geologic insight and authoritative segmentations of geologic units. The information system can examine many images
quickly, quantify features with accuracy and consistency, and perform exhaustive statistical analysis to search for unexpected patterns
and anomalies. In this sense the roles of the geologist and the automated mapping system are strongly complementary. The mapping
system automates some of geologists' more arduous tasks, freeing them to focus on interpretation.
| |
| Improving Swift (contd) (PI: Levison, Harold,
Southwest Research Institute
) | |
| We propose to consolidate and extend the work begun under previous AISR grant
entitled "Improving Swift." Swift, developed in large part and supported by
the PI, is a state-of-the-art software package for performing numerical
simulations of the dynamic evolution of planetary systems.Under our previous
grant we have made a significant number of fundamental improvements to Swift (
we call the improved package "Swifter"), and we have developed an associated
graphical data analysis package (called "SWIFTVis"). Both Swifter and
SWIFTVis, however, require further testing before we can confidently release
them to the public. Therefore, we will perform an extensive suite of build
level tests in order to facilitate a public release of Swifter and SWIFTVis as
soon as possible. Also, we will further develop tree-based force calculation
algorithm for use with the numerical integrators of Swifter, thereby
enabling far larger and more encompassing simulations than have been
heretofore possible. Finally, we will develop a new integration algorithm
that promises to enable fundamentally new research in the field of planetary
system formation, in particular the formation of planets around stars in
their birth clusters.
| |
| Automated Identification and Characterization of Landforms on Mars (PI: Stepinski, Tomasz,
USRI/LPI
) | |
| We propose development, implementation, testing, and application of novel methods for automated identification and analysis of selected landforms on Mars. Because of the large and increasing volume of Martian data such automated techniques are essential for thorough data analysis. In particular, we will implement algorithms for an identification and characterization of Martian impact craters and valley networks. In a major break with the past attempts to automate identification of landforms on planetary surface, the proposed algorithms are based on digital elevation data instead of imagery data. Using a fusion of methods from the fields of digital terrain analysis, data mining, and computer vision, we will produce production grade, fully autonomous algorithms. The crater identification algorithm will be used to construct a catalog of Martian craters that, in addition to craters coordinates, lists all their physical parameters. Such catalog would be an invaluable resource for a range of Martian studies. The valley networks mapping algorithm will be used to map drainage systems and to produce continuous maps of drainage density. These are important resources in the studies of past Martian climate. The proposal offers proof-of-concept results for these two algorithms. In addition, we propose to improve on our recently developed method for a general, unsupervised landform classifier, which subdivides Martian landscape into constituent landforms. We will employ the technique of semi-supervised clustering to increase the accuracy of the original classifier. We expect that the new classifier will provide, among other benefits, an alternative method for crater identification and analysis. The methods developed in the proposed research are applicable to other bodies in the Solar System, such as the Earth, the Moon, and Mercury, for which the topography data is available, or is about to become available. By automating analysis of some of the most interesting landforms on Mars, the proposed research increases efficiency of the Science Mission Directorate research, which is one of the goals of the AISRP. It is highly relevant to the NASA objective of thorough study of planet Mars.
| |
| Integration of Orbital, Descent and Ground Imagery for Topographic Capability Analysis in Mars Landed Missions (PI: Li, Rongxing,
Ohio State U
) | |
| The goal of the proposed research is to develop advanced geo-information technology to integrate orbital, descent, and ground imagery for topographic capability analysis in Mars landed missions. Objectives of the project are to: (1) Investigate how to integrate the orbital, descent and ground imagery within a true 3D integrated geometric model, (2) Estimate the precision of topographic features mapped on the Martian surface, and (3) Develop innovative applied information methods to register orbital, descent and ground imagery through distinctive landmarks, and to develop methods for automatic mapping of large terrain features such as major craters, hollows, and hills.
Systematic assessment of topographic mapping capabilities of orbital, descent and ground imagery will be conducted based on photogrammetric bundle adjustment and statistical analysis. The current bundle-adjustment technology, which has been used during MER (Mars Exploration Rover) mission operations for MER landing-site mapping and rover localization using ground images, will be extended to include orbital and descent images within an integrated geometric model. Using an extended bundle adjustment, the positions and orientations of all three types of images will be adjusted in this integrated image network to achieve the best possible accuracy. Topographic products, such as digital terrain models (DTMs), orthophotos, and 3D dynamic models (VRML), will be generated after bundle adjustment of the image network. These products are very important for topographic and geological characterization of landing sites, for crater mechanics modeling, and for other scientific research in future Mars landed missions.
The data to be used to verify the developed technology include MOC NA imagery (cPROTO, 0.5-m resolution), Mars Express HRSC stereo imagery (10-m resolution), MOLA original profiles and DTMs (Digital Terrain Models), descent (DIMES) images, and MER Pancam and Navcam stereo images taken at the two landing sites of the MER mission.
The proposed research goes beyond existing efforts because it integrates orbital, descent and ground imagery in a new comprehensive geometric model and facilitates high-precision topographic mapping. The developed methods can be used in future missions for regional topographic mapping to support landing-site selection, for topographic characterization before landing, and for mission planning and surface operations at an extended landing site.
| |
| Intelligent Sensor Network Study of Dust Devils (PI: Lorenz, Ralph,
Johns Hopkins University Applied Physics Lab
) | |
| We propose to perform a novel field study of terrestrial dust devils in Arizona using initially an array, evolving into an intelligent network, of smart sensor platforms. Dust devils are a dynamic feature of the Martian environment, playing an important part in climate via dust loading, and can significantly affect surface operations by sweeping dust off solar panels as well as possible electrostatic discharge effects. Even on Earth, dust devils are implicated in ~100 light aviation accidents in the last 15 years and thus can be considered a genuine hazard. A systematic study, which without the application of intelligent systems would be a labor-intensive effort, is therefore suggested. An array of rapid-sampling sensor platforms can generate datasets on the formation, structure and evolution of dust devils that can be compared with mesoscale numerical models. After initial data acquisition, algorithms will be developed by which the field network can autonomously identify the presence of dust devils and generate reports. Intelligent sensing yields orders of magnitude reduction in data volume.
| |
| Adaptive Algorithms For Optimal Classification and Compression of Hyperspectral Images (PI: Bose, Tamal,
Utah State U
) | |
| Hyperspectral images are obtained from large focal plane spectrometers with an array of multiple sensors that are arranged in an NxN matrix. Each sensor yields an Mx1 spectrum, where M is the number of image bands, typically between several dozens to several hundreds. Spectral and hyperspectral images are used today in most space science and earth science missions because of the richness of information they capture about the compositional properties of surface materials. Since hyperspectral images constitute very large data sets, they need to be compressed for storage and transmission. Classification of spectral species is the primary task in the analysis of hyperspectral imagery, with the goal of distinguishing as many species as possible, as accurately as possible. Therefore, a measure of good compression should be based on how well it preserves class distinctions. This, in general, does not correspond to a minimum distortion measure.
The goal of this project is "classification-driven compression" of hyperspectral images. That is, we propose to design and implement compression algorithms that will be optimized for efficient classification. The compression algorithms will be recursively optimized as a function of certain classification (or unsupervised clustering) metrics. Performance metrics will be defined for the classification schemes and some others for the compression algorithms. A combination of these metrics will be used to design a cost function, which in turn will be optimized to update the compression algorithms. In other words, the performance of the classification algorithms will dictate the real-time adaptation of the digital filters used in the compression schemes. This is a new concept, where the filters in the compression algorithms, instead of acting independently, form a closed loop coupled system with the classification (or clustering) algorithms. The algorithms will be designed for real-time on-board processing. Therefore, we seek to develop fast filtering algorithms with the lowest possible computational complexity. This work will change the way we look at and understand hyperspectral data. Since we will measure the quality of the compression by the retention capability of the distinction among spectral species, for any required compression ratio, we will much more successfully preserve the meaning of the data inherent in and specific to spectral images, than distortion based compressions. The proposed work was inspired by and will build on the results of two previous AISRP grant projects.
All space science missions that carry spectral imagers can significantly benefit from the results of this project in scenarios where lossy data compression is needed at compression ratios as high as can be achieved without loss of relevant spectral information. Because of our approach to compression, we will be able to determine the optimal compression ratio in an adaptive manner. The fast parallel implementation of the compression and classification algorithms, makes the combined system described in this proposal an intelligent, real-time on-board data understanding, compression and classification machine that learns on-chip and adapts continuously to new circumstances as desired, modifying the compression scheme to best suit a given environment.
| |
| A neural map view of planetary spectral images for precision data mining and rapid resource identifi (PI: Merenyi, Erzsebet,
Rice University
) | |
| This project follows up on a current three-year AISRP grant, NAG5-10432, which will end 8/2004. It addresses a pressing need for rapid yet intelligent analysis of voluminous multi- and hyperspectral images in order to extract key data and generate knowledge. Spectral imaging plays a leading role in remote identification of surface materials of Earth (Landsat, AVIRIS, Hyperion), Mars (Pathfinder, MGS, MER), the Jovian system (Galileo NIMS), the Saturnian system (Cassini VIMS) and other solar system bodies. Hyperspectral sensors, in particular, enable detailed identification through the complexity of signatures measured in hundreds of narrow spectral bands. The challenge in automated and fast (real time, on-board) interpretation of these huge images calls for massively par-allel algorithms, as well as it requires sophisticated algorithms for optimal knowledge extraction. Properly utilized, Artificial Neural Networks (ANNs) can provide both. The current project engineered ANNs, specifically Self-Organizing Maps (SOMs) and their hybrids for efficient and sophisticated clustering and classification of spectral images, developing custom modules supported by commercially available components. Based on some of the latest theoretical research on SOMs, the tools we developed, jointly with European experts, are powerful for distinguishing a large number of spectral classes and for the discovery of ''interesting'' but uncommon and spatially very small classes. We use information theoretically principled SOM approaches, which increases power and confidence in autonomous data mining. We demonstrated the effectiveness and high quality of data analysis on sample IMP spectral images, Cassini VIMS Jupiter fly-by im-agery, AVIRIS and other data representing typical challenges in NASAís missions. We propose to advance these computational intelligence capabilities in three ways: 1) We will add significantly new theoretical strength to information extraction modules. 2) The software, HYPEREYE, will be made transferable to other users through high-level graphic interface, augmented software design, tutorials and wrapping, opening an important phase of technology infusion that will take recent and future developments into the user community. 3) We will directly participate, using our methods and software, in analyses of spectral images forthcoming from the Mars Exploration Rovers and Cas-sini VIMS Saturn orbital tour, and (pending its funding) a Pluto/icy satellites spectral analysis project. The ëneuralí core of our software is already suitable for implementation in high-speed massively parallel hardware (which could be an on-board analysis capability), as it was one of the original objectives of our work. We are pursuing that line of development outside of this project proposal and, if successful, we anticipate using the hardware to support this work as well.
| |
| Application of Machine Learning Technology to Martian Geology (PI: Wang, Ruye,
Harvey Mudd College
) | |
| New and more powerful machine learning technologies have always been sought to effectively analyze multi/hyperspectral remote sensing data, which have been rapidly accumulating by the ever-increasing number of spacecrafts launched by NASA. In particular, algorithms for more sensitive detection and accurate classification are necessary to help identify, map, and characterize materials with subtle spectral differences such as a diverse suite of rock types that may exist on the surface of Earth or other planets such as Mars. With ever-increasing data from ongoing and near-term missions, such tested and proven techniques will greatly benefit the planetary community, as traditional means of data analysis may not produce the required sensitivities for remote detection and accurate classification of various surface materials. We propose to develop and test an intelligent system composed of a suite of the latest cutting-edge algorithms for the detection and classification of surface materials of weak and/or similar spectral signatures that are otherwise difficult to detect and distinguish by the existing conventional methods. The algorithms include, for example, the kernel-based methods, such as support vector machines (SVMs), which have demonstrated their superiority over many conventional classifiers in terms of classification accuracy and the independent component analysis (ICA) methods, which is capable of unmixing the observed signals to find the independent components with minimum prior knowledge and additional information, and is therefore a powerful tool to separate mixed materials contained in remote sensing image data. While such methods have already found applications in different scientific fields, here we propose to develop an intelligent system by coupling new methods and algorithms with the strength of these methods for the purpose of addressing a variety of space science related problems. In particular, we will use the system to more accurately assess the rock types in the ancient mountain ranges of Mars, Thaumasia highlands and Coprates rise, which includes determining whether the ranges consist of more silicic-rich, mountain building rock. The identification of such a suite of rocks would ultimately lead to an improved understanding of the geological evolution of early Mars.
| |
| Advanced Visualization in Solar System Exploration and Research (ADVISER) (PI: Head, James,
Brown University
) | |
| We propose 1) to advance space science knowledge, exploration capabilities, and teaching and outreach, and 2) to research advanced visualization tools for space science and education, through ADVISER (Advanced Visualization in Solar System Exploration and Research), a problem solving environment (PSE) for planetary geosciences that will integrate and extend the state-of-the-art in hardware and software technologies. Our three-pronged science approach (Mars polar evolution, tropical glaciers, and the Noachian hydrological cycle) builds on our central strengths in planetary geoscience, as well as supporting NASA Goals and Objectives and the basic NASA research and exploration theme for Mars ''Follow the Water'', and thus will provide scientifically credible direction for tool research and scientific discovery. Geologists explore the Earth primarily through field work and analysis of the geological record at various points on the surface, integrating these individual points of understanding using more synoptic analysis. Planetary geoscientists commonly work in the reverse order, starting from synoptic and orbital data and working back down toward the surface. New developments in advanced visualization, and immersive virtual reality environments have created the ability to place the geoscientist back down on and near the surface of the planet and to regain the perspective that is the foundation for the understanding of relationships necessary to unlock the history of the planets. ADVISER is a set of tools that provide the planetary geoscientist with the capability to operate and analyze data as if they were on or near the surface of a planet. The ADVISER PSE has four basic parts: 1) Geoscientist on the Surface: Visualization capabilities that enable the placement of the geoscientist into the surface environment through immersive virtual reality (IVR) and related desktop capabilities using topographic data and surface rendering programs. 2) Importation and Visualization of Multiple Data Sets: On-demand importation, co-registration and overlay of relevant image format data sets to enhance the ability of the geoscientist to correlate and interpret data for scientific analysis. 3) Field Kit Development: A menu that has the capabilities commonly carried out by the geologist in the field using things like the Brunton compass, altimeter, etc. 4) Ancillary Virtual Field Instrument Development: a) Virtual Photography: Permits documentation of individual images or video streams. b) Virtual GPS: Permits exact location at any point in the analysis and tracking and storage of traverses. c) PDA: Equivalent of the field notebook and permits the investigator to record a host of information for synthesis and publication. Together, these products will contribute significantly to both the scientific and pedagogical goals of the AISR program.
| |
| Analyzing Science Operations for the Search for Life as part of a Multi-Year Robotic Campaign to Exp (PI: Thomas, Geb,
University of Iowa
) | |
| In the austral spring (September-October) of both 2004 and 2005, a multi-institutional, multi-national effort led by Carnegie Mellon University will conduct long-duration robotic explorations of the Atacama Desert in Chile to develop technologies and methods for upcoming NASA missions that will search for evidence of past life on Mars. The project proposed herein will leverage this large-scale effort to measure and improve the effectiveness of robotic science operations. This research will extend current analysis of rover-mediated geology to rover-mediated habitat characterization. The work will continue to emphasize the effects of different data collection and display techniques on the science teamís conclusions. The principal hypothesis this research is that the quality and reliability of scientific conclusions regarding past or present life in arid environments is dependent on the type of evidence collected by the rover, the scientistsí data analysis techniques, the processes used by the scientistsí to form and share hypotheses and conclusions, and the science operations software. The principal objectives of this research, each specifically associated with AISR program objectives, are to: 1) reduce mission development time by analyzing how scientists characterize a habitat, 2) reduce mission development risk by identifying mission-critical and problematic analysis tasks, 3) increase science return from the data by analyzing long-traverse science collection strategies, and 4) increase data return by refining the science interface to improve analysis effectiveness. The proposed research will analyze the processes used by astrobiologists and geologists when searching for signs of life in a Mars-like environment. Our previous and ongoing work with robotic geology has successfully characterized limitations in scientific interpretations caused by rover sensors, differences in scientists' interpretations, and limitations of the science interface. These limitations were identified and studied using perceptual experiments in which scientists analyzed sample images and physical specimens. In addition, transcripts of scientists participating in a simulated rover field experiments have been examined to further understand these limitations. The proposed project will quantify analyst and instrument limitations that could affect the success of future missions in the search for life on Mars and will develop mitigating strategies to avoid inappropriate conclusions regarding the presence of life on Mars.
| |
| Neural Net Spectral Analysis of Icy Volatiles in the Solar System (PI: Young, Eliot,
Southwest Research Institute
) | |
| We will train and apply a two-stage neural net algorithm to spectra of satellites (Rhea, Mimas, Dione, Europa, Ganymede, Callisto, Titania, Umbriel, Ariel, Oberon, Miranda, Triton and Charon), Centaurs (Chariklo and others) and Pluto all of which have surface water ice or methane ice. These two ices have IR spectral features that are diagnostic of composition, grain size, surface temperature and crystalline state, but the interpretation typically depends on subtle spectral features that defy many spectral classification schemes. The proven neural net algorithm we will use has a self organizing map as a preprocessor to help robustly extract spectral information. This two-tier neural net has already achieved excellent results in several diverse palentary applications.
| |
| A Numerical Simulation Tool for Planetary Subsurface Radar (PI: Cummer, Steven,
Duke University
) | |
| Exciting recent discoveries in planetary science have made subsurface exploration instruments one of the top priorities on future planetary science missions. Remote sensing techniques, specifically ground penetrating radar (GPR), could be very effective in the critically important search for water. GPR instruments are planned on many planetary missions in the coming decade. But there are many challenges to successfully operating planetary GPR instruments and interpreting the data. These challenges create a compelling need for meaningful quantitative analysis of environmental impacts on instrument operation and of GPR data interpretation. We propose to develop an accurate and general numerical modeling tool to simulate the planetary GPR problem. This tool would have substantial value to the planetary science community as a means for realistic numerical experimentation of an environment completely unavailable for actual field experiments. This code would find immediate use in quantifying the capabilities of planned GPR instruments in light of major environmental unknowns, and providing realistic signals on which data processing schemes can be tested. In the future use, the proposed model would be valuable in bounding the interpretation of any data returned from these instruments. The code development addresses a compelling need in the planetary science community and is targeted directly at AISR program goals.
| |
| Adaptation & Use of OpenGIS Web Technologies for Multi-Disciplinary Access To Planetary Data (PI: Dobinson, Elaine,
Jet Propulsion Laboratory
) | |
| We propose to adapt the OpenGIS standards and technologies for the access, processing, and display of Earth geospatial data to the lunar and planetary domains. In particular, we will implement prototype WMS and WCS servers to present high level data products derived from the Mars and lunar datasets archived within the Planetary Data System (PDS). The specific information technologies used for this work are the rapidly developing and well-supported OpenGIS methodologies for Earth data. We will extend and generalize these standards for terrestrial planetary bodies, using Mars and the Moon as both subject of the prototype implementations and the exemplar for standards extensions. GIS software is being increasingly used by the planetary, and, in particular, the planetary geologic mapping communities. To our knowledge, however, this would be the first concerted effort to build a general data delivery service that would feed and augment the use of GIS commercial software for data processing and analysis. The basic motivation is that, while the World Wide Web has brought a revolution to the PDS in terms of ease of access, the accessible products remain relatively low level. Planetary and lunar investigators will benefit from a unified methodology for accessing higher-level products that serve as substrate, background, and currently known detail pertinent to their ongoing work. The thrust of this effort is to capitalize on a large body of existing technology and bring it to bear for the benefit of the planetary community. To be successful we will need to determine the minimum changes to that existing technology needed for the adaptation process. The central challenge is to craft a coherent set of changes that are comprehensive enough to apply to planetary bodies in general, while circumscribed enough to keep the work within the prescribed scope. This work directly pursues a major objective of the AISR call: ìIncrease science or educational return from the data through, for example, advanced knowledge discovery, data synthesis and data presentation methodologies.î Since we anticipate our results will have the most impact relative to the study of the Moon and the terrestrial planets, we connect most directly to the following NASA Goals, Science Objectives, and Research Focus Areas: Goal II, Solar System Exploration Theme, RFA 1(c); Goal II, Solar System Exploration Theme, RFA 4(c); and, Goal II, Solar System Exploration Theme, RFA 6(a). The nature of our work will lead to the broad publication of global multi-phenomenological maps of lunar and planetary bodies. These maps will be hierarchical and afford both synoptic and high-resolution views of each body published. The multiple views will make clear what differences exist, and, because they present the different phenomena as a composite, this capability should provide a basis to better understand planetary processes. We have assembled a highly expert and broadly based team for these tasks. With balance in both Information Technology and Planetary Science, we bring many years of direct experience with the PDS and with the scientific mapping of planetary bodies. Most pertinently, we have successfully brought OGC specification and technology to the management and publication of massive geospatial datasets.
| |
| Multi-Scale Atmospheric Numerical Modeling and Data Assimilation for Planetary Applications - With a Focus on Mars (PI: Richardson, Mark,
California Institute of Technology
) | |
| The goal of this proposal is the development of the first flexible, multi-scale numerical, dynamical model of planetary atmospheres - and specifically the Martian atmosphere. The model will provide a single tool for study of atmospheric motions on scales from the micro (tens to hundreds of meters) to global; it will provide full flexibility in spatial grid construction; and will provide a framework for Martian data assimilation and weather prediction. The project will take advantage of the new Weather Research and Forecast (WRF) modeling system developed by a consortium of US universities and government organizations - NCAR, NOAA and the DoD - and will thus track (and in some key areas lead) the state-of-the-art in terrestrial model design. This project will see the infusion of a number of modern software coding practices into planetary/Martian atmospheric modeling, including extensive modularity, from-the-ground-up use of fast and compact Fortran 90/95, and explicit MPI parallelization. Support is requested to allow conversion of this community model to allow for generalized planetary use (and for Martian conditions in particular) and to undertake changes in the dynamical core to allow it to run as a global model. By leveraging substantial investment in model development by other agencies, this effort will provide a very high return for NASA investment ñ as well as benefiting those other agencies. We intend to feed the model improvements from this project back into the public-domain WRF system, making it available for all NASA planetary and Earth systems science researchers.
| |
| Data-Centric Analysis of Science Return for Human-Directed Robotic Geology (PI: Thomas, Geb,
University of Iowa
) | |
| In the austral spring (September-October) of both 2004 and 2005, a multi-institutional, multi-national effort led by Carnegie Mellon University will conduct long-duration robotic explorations of the Atacama Desert in Chile to develop technologies and methods for upcoming NASA missions that will search for evidence of past life on Mars. The project proposed herein will leverage this large-scale effort to measure and improve the effectiveness of robotic science operations. This research will extend current analysis of rover-mediated geology to rover-mediated habitat characterization. The work will continue to emphasize the effects of different data collection and display techniques on the science team�s conclusions. The principal hypothesis this research is that the quality and reliability of scientific conclusions regarding past or present life in arid environments is dependent on the type of evidence collected by the rover, the scientists� data analysis techniques, the processes used by the scientists� to form and share hypotheses and conclusions, and the science operations software. The principal objectives of this research, each specifically associated with AISR program objectives, are to: 1) reduce mission development time by analyzing how scientists characterize a habitat, 2) reduce mission development risk by identifying mission-critical and problematic analysis tasks, 3) increase science return from the data by analyzing long-traverse science collection strategies, and 4) increase data return by refining the science interface to improve analysis effectiveness. The proposed research will analyze the processes used by astrobiologists and geologists when searching for signs of life in a Mars-like environment. Our previous and ongoing work with robotic geology has successfully characterized limitations in scientific interpretations caused by rover sensors, differences in scientists' interpretations, and limitations of the science interface. These limitations were identified and studied using perceptual experiments in which scientists analyzed sample images and physical specimens. In addition, transcripts of scientists participating in a simulated rover field experiments have been examined to further understand these limitations. The proposed project will quantify analyst and instrument limitations that could affect the success of future missions in the search for life on Mars and will develop mitigating strategies to avoid inappropriate conclusions regarding the presence of life on Mars.
| |
| Autonomous Mineral Detectors for Mars Rovers and Landers (PI: Gilmore, Martha S.,
Wesleyan University
) | |
| We will design software that will provide instrumented rovers the ability to select and analyze spectroscopic data, characterize a landing site and test geological hypotheses autonomously. Our approach is to combine supervised methods for focused mineral detection and broader mineral classification with an unsupervised clusterer that identified unusual or previously unencountered mineral classes. We will 1) improve our current carbonate detector by validating its sensitivity and performing more detailed modeling of nonlinear mixture behavior observed in intimate mineral mixtures, 2) identify other mineral classes of interest, particularly those known to exist on Mars and be indicators of life on Earth, and design detectors for those minerals, and 3) develop a novelty detector that can identify new minerals in a series of observations. The algorithms will be trained using synthetic data from spectral libraries and tested in real data collected in the field.
| |
| Improving Swift (PI: Levison, Harold,
Southwest Research Institute
) | |
| Understanding the formation of planets is one of the most fundamental problems facing astronomy today. It is also an area of great current interest since NASA has determined that understanding the origin and evolution of our solar system and planetary systems around other stars is one of its fundamental scientific goals. Perhaps the best method for understanding how the solar system formed is by performing large-scale numerical experiments. Such simulations require computer codes that can integrate the orbits of a large number of bodies for, in many cases, timescales approaching the age of the solar system. Over the last 10 years, PI Levison and CoI Duncan have developed a software package, known as Swift, capable of performing these types of simulations. Swift quickly became an industry standard for the planetary community, and is used all over the world. Here we propose to make several fundamental improvements to Swift that would significantly enhance its utility to the planetary community. We plan to: (1) employ more sophisticated methods for calculation the forces between objects, (2) parallelize parts of the code to take advantage of the current popularity of Beowulf clusters, and (3) rewrite the code and thereby utilize more efficient data structures.
| |
| Capturing Planetary Science Data from NASA Supported Research (PI: Sykes, Mark V.,
University of Arizona
) | |
| A robust web interface will be built, creating the capability for hundreds of NASA PIs to submit planetary data and ancillary files to the NASA Planetary Data System, an archive which currently
focuses on the ingestion of large flight project databases. This interface will be transparent, requiring no knowledge of PDS specifics by the submitter. Utilizing information from the submitter, it will generate all necessary PDS internal files and formats which in turn will be automatically run through currently existing PDS standards validation software. The interface design will take advantage of common data structures across science enterprises and take advantage of the platform independence of the World Wide Web and increasing transfer rates of the Internet. This interface will allow for the rapid availability of a large amount of diverse data generated by OSS research programs, allowing for greatly increased scientific return from those programs. Prototype and final interface designs and pipelines will be reviewed by non-PDS researchers and PDS scientists and data engineers to ensure usability, efficiency, and reliability before final delivery to the PDS. The PDS Small Bodies
Discipline Node (managed by the Co-I, Dr. Michael F. A'Hearn) has committed to testbedding this interface as a precursor to broader implementation across all relevant PDS Discipline Nodes. Because
most of these data structures are also common to astrophysics data sets, this will also have applicability to astrophysics data archives.
| |
|
|
|
|
|
