1 / 33

NASA Sensor Web Activities

NASA Sensor Web Activities. CEOS WGISS-23 Hanoi, Vietnam. Martha Maiden Program Executive Earth Science Data Systems NASA Headquarters. Overview of Presentation. Sensor Web Research at NASA NASA Sensor Web Vision and Definition

harlow
Télécharger la présentation

NASA Sensor Web Activities

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. NASA Sensor Web Activities CEOS WGISS-23 Hanoi, Vietnam Martha Maiden Program Executive Earth Science Data Systems NASA Headquarters

  2. Overview of Presentation • Sensor Web Research at NASA • NASA Sensor Web Vision and Definition • NASA’s EO-1 Satellite Use For Sensor Web Technology Demonstration • NASA A-Train Data Depot

  3. Advanced Information System TechnologySensor Web Research • The goal of the Earth Science Technology Office (ESTO) Advanced Information System Technology (AIST) Program is to develop and mature improved information system technology and provide risk reduction for NASA’s Earth Science Data Systems. • NASA’s AIST Program dedicated a solicitation to Sensor Web in 2005

  4. Sensor Webs for Earth Science - NASA Perspective A Vision for NASA Sensor Webs for Earth Science: On-demand sensing of a broad array of environmental and ecological phenomena across a wide range of spatial and temporal scales, from a heterogeneous suite of sensors both in-situ and in orbit. Sensor webs will be dynamically organized to collect data, extract information from them, accept input from other sensor / forecast / tasking systems, interact with the environment based on what they detect or are tasked to perform, and communicate observations and results in real time. • A Sensor Web is a coordinated observation infrastructure composed of a distributed collection of resources - e.g., sensors, platforms, models, communications infrastructure - that can collectively behave as a single, autonomous, task-able, dynamically adaptive and reconfigurable observing system that provides raw and processed data, along with associated meta-data, via a set of standards-based service-oriented interfaces. ( FromAIST meeting with NASA Sensor Web PI’s, ~40 participants discussing concepts, architectures, features and benefits of sensor webs)

  5. Key Capabilities Implemented to Enable EO-1 Sensor Webs

  6. Six ESTO AIST 05-Funded Studies Used in the EO-1 Sensor Web Demo • Dynamically Linking Sensor Webs with Earth System Models: PI - Liping Di/George Mason Univ. Methods to discover science products and invoke algorithm workflows automatically • Increasing TRL of SensorML: PI - Mike Botts/Univ. Alabama, Huntsville Use SensorML for discovery and invoking algorithm workflows • Sensor-Analysis-Model Interoperability: PI - Stefan Falke/Northrop-Grumman/Washington Univ. St. Louis Standard model interfaces to drive sensor webs • Sensor Web Dynamic Replanning: PI - Steve Kolitz/Draper Labs Decision support systems for sensor webs Cloud screening for optimizing tasking of satellite assets • Using Intelligent Agents to Form a Sensor Web Base for Autonomous Operations: PI - K. Witt/WVHTF Help implement SensorML use to describe sensor capabilities for discovery • Volcano Sensor Web: PI - Ashley Davies/NASA-JPL Detect and image volcanoes autonomously with EO-1

  7. Various EO-1 Sensor Web Experiments Conducted

  8. Making EO-1 Discoverable on the Internet

  9. OGCSensor Web Enablement -2- • Sensor Web Enablement Framework - Schema • SensorML – models and schema for describing sensor characteristics (geolocation, response) • Observation & Measurement – models and schema for encoding sensor observations

  10. OGC Sensor Web Enablement -3- • Sensor Web Enablement Framework – Services • Sensor Observation Service – access sensor information (SensorML) and sensor observations (O&M) • Sensor Planning Service – task sensors or sensor systems • Web Alert Service – asynchronous notification of sensor events (tasks, observation of phenomena) • Sensor Registries – discovery of sensors and sensor data

  11. Sensor Modeling Language • XML based • Provides general sensor information to support data discovery • Supports processing and analysis of sensor data • Supports geo-locations of sensor data • Provides performance characteristics (accuracy, thresholds, etc.) • Archive fundamental properties and assumptions regarding sensors • Can apply to any sensor whether in-situ or remote • Facilitates “plug and play” and interoperability between sensors • Especially useful for heterogeneous sets of sensors and rapid integration of new sensors From---- http://vast.nsstc.uah.edu/SensorML/

  12. A General Framework and System Prototypes for the Self-Adaptive Earth Predictive Systems (SEPS)--Dynamically Coupling Sensor Web with Earth System Models PI: Liping Di, GMU, Co-I: James Smiths, GSFC, David Lary, UMBC Objective Scientists from GMU, GSFC, and UBMC will collaborate to 1) develop a general Self-Adaptive Earth Predictive Systems (SEPS) framework for dynamic, interoperable coupling between Earth System Models (ESMs) and Earth Observing (EO) sensor web and data systems, based on open, consensus-based standards; 2) implement and deploy the framework and plug in diverse sensors and data systems to demonstrate the plug-in-EO-and-play capability; and 3) prototype a Bird-Migration-Model-to-aid-avian-influenza-prediction SEPS and an atmospheric chemistry composition SEPS using this framework, to demonstrate the framework’s plug-in-ESM-and-play capability and its applicability as a common infrastructure for supporting the focus areas of NASA research. DDRS—Data Discovery and Retrieval Services, PIAS—Pre-processing, Integration, and Assimilation, SGMS—Science Goal Monitoring Service, DSPS-Data and Senor Planning Services, CENS—Coordination and Events Notification Services SEPS and the SEPS Framework • Accomplishments • Demonstrated the feasibility of adopting the self-adaptive concept for the coupling of models with sensors. • Enabled the targeted sensor observation through science goal monitoring controlled by model predictions. • Demonstrated through the initial work that the framework can work easily with different Web-ready sensors, such as NASA Ames UAV and the EO-1 (Note: OGC Sensor Web Enablement Interfaces have been implemented on Ames UAV and EO-1 to make them sensor-web ready). Schedule and deliverables • Project startup 09/2006 • OGC Sensor Web Demo 11/2006 • ESIP federation sensor web demo 07/2007 • Completion of the feed-back segment 12/2007 • Completion of the feed-forward segment 12/2008 • Bird-flu SEPS prototypes demo 04/2009 • Atmospheric chemistry composition SEPS demo 08/2009 • All software developed will be freely available to NASA and its partners.

  13. Increasing the Technology Readiness of SensorML for Sensor Webs PI: Mike Botts, UAH Objective We will reduce the current challenges involved in implementing and utilizing SensorML by providing a collection of Open Source tools for creating, viewing, validating, mining, and executing SensorML processes. We will also demonstrate the application of these tools, and indeed the application of SensorML, in an end-to-end scenario of relevance to NASA’s Earth Science community, including the derivation of SensorML documents by the initial sensor team, the configuration of OGC sensor web services, the development of product algorithms by research scientists, and the ultimate discovery and application of SensorML within the end user’s Decision Support Tools. SensorML Process Editor Approach Key Milestones Some of the proposed software will be brought to a TRL of 6 since they will be fully used at critical points throughout the framework. These include: - SensorML Document Parser (target TRL 6/7) - SensorML Document Validator (target TRL 6/7) - SensorML Processing Engine (target TRL 6/7) - SensorML Data Miner (target TRL 6/7) Some tools will be developed and brought to a TRL of 4 since they will remain experimental. - SensorML Viewers and Editors (target TRL 5) - SensorML Distributed Processing Engine (target TRL 5) • SensorML Parser/Validator, Execution Engine (V1) 3/2007 • SensorML Viewer/Editor (beta) 3/2007 • SensorML Miner & Registry (V1) 9/2007 • SensorML Process Execution engine (V2) 3/2008 • SensorML Editor & Viewer (V1) 3/2008 • Sensor Descriptions/Geolocation/Web Services 9/2008 • Advanced Product Processes 3/2009 • Decision Support Tool / Advanced Tools 3/2009 • End-to-End Demonstration 9/2009 Co-I’s/Partners • Alexandre Robin, Anthony Cook /UAH TRLin = 2-4

  14. Using Intelligent Agents to Form a Sensor Web for Autonomous Mission Operations PI: Kenneth Witt, West Virginia High Technology Consortium Foundation Objective We will develop an architecture which shifts sensor web control to a distributed set of intelligent agents versus a centrally controlled architecture. Constellation missions introduce levels of complexity that are not easily maintained by a central management activity. A network of intelligent agents reduces management requirements by making use of model based system prediction, and autonomic model/agent collaboration. The proposed architecture incorporates agents distributed throughout the operational environment that monitor and manage spacecraft systems and self-manage the sensor web system via peer-to-peer collaboration. The intelligent agents are mobile and thus will be able to traverse between on-orbit and ground based systems. GMSEC and CFE/S features include Plug-and-Play Components, and Standard Messages implementing a software Information Bus. Approach Key Milestones Our team will develop and integrate these technologies: Model Based Operations, Intelligent Agents, Software Bus Architectures, and Sensor Webs. EO-1 and ST-5 have successfully demonstrated that model based operations can support autonomous control of a satellite mission. The next step is to connect the autonomous operations that take place on the platform to those happening on the ground. • Initial Architecture Document Feb(April)/2007 • Bus-Bus bridge and CHIPS demonstration April/2007 • Mobile agent demonstration Nov/2007 • Basic framework report capability July/2008 • Updated architecture documentation August/2008 • Final architecture document June/2009 • Comprehensive demonstration August/2009 Co-I’s/Partners • Al Underbrink / Sentar Inc. • Daniel Mandl / GSFC TRLin = 3

  15. Science Model Driven Autonomous Sensor Web PI: Ashley Davies, JPL Objective To maximize science data return and optimize asset and resource use of an existing sensor web by including volcanic process models in the control loop. We will modify an existing sensor web that has a simple trigger-reaction mode, to one that uses a volcanic process model to guide the reaction. For example: a ground sensor detects increasing activity, causing the sensor web to seek additional key data as input for a model of a volcanic process to determine volcano state. This effort will integrate automated retasking and science process modeling to enable true science-driven sensor web operations. Data flow of proposed prototype Model-based Sensor Web. Approach Key Milestones • Define the “State of the Volcano” and track this state using SensorML, with integration of an eruption process model, and with automated data processing and asset re-tasking. • Demonstrate an autonomous ‘closed loop’ of information transfer from trigger event to processing through the sensor web hub at JPL, spacecraft observation, data analysis, and back to the trigger origin (to volcanologists in the field). • Complete “State of the Volcano” definitions 03/07 • Complete Sensor Web design 04/07 • SensorML coding complete 09/07 • Field testing and verification 11/07 • Demonstrate operational system 12/07 Co-I’s/Partners Rebecca Castaňo & Steve Chien (JPL); Robert Wright (U. Hawai’i), Philip Kyle (New Mexico Tech.), Thomas Doggett (ASU), Felipe Ip (U. Arizona) TRLin = 3 4/07

  16. Virtual Constellations - What Sensor Web Capabilities Needed? • It appears there are 3 major implementation parts of the sensor web: 1. Collecting sensor web data from a number of related instruments 2. Combining the data to create information for user or sensor feedback 3. Communicating information back to sensors, near real time, for information sharing and feedback reaction (i.e., alter sensor behavior; e.g. alter measuring schema) • NASA ACCESS Project “A-Train Data Depot” is addressing parts 1 and 2. The Depot is can be found online at the following URL: http://disc.gsfc.nasa.gov/atdd/index.shtml • Possible future work can address part 3.

  17. NASA’s A-Train:Precurser to Virtual Constellation • NASA’s A-Train was created “after the fact” of launch, by creating a closely-spaced track in polar orbit at about 1:30 pm equator crossing time. NASA’s Aqua, CALIPSO, Cloudsat, and Aura are joined by French PARASOL satellite (between Cloudsat and Aura).

  18. Cooperative Institute for Research in the Atmosphere (CSU): Cloudsat Atmospheric Science Data Center (LaRC): CALIPSO, TES, CERES Level 1 and Atmospheres Data System (GSFC): MODIS A-Train Data Depot The primary sensor web node that: Gathers heterogeneous sensor web measurements; Prepares measurements for intercomparison research, and; Provides access to measurements for multi-sensor data analysis and future inter-sensor communications. Data from A-Train sensor web nodes Data Centers Goddard Earth Sciences Data and Information Services Center (GES DISC) Atmospheric Composition DISC: AIRS, MLS, OMI, HIRDLS Remote access to A-Train specific data A-Train Data Depot • S4PM inter-sensor data processing: • Data co-registration • Coincident subsetting • -Data Fusing Data Search Inter-sensor information Data Exploration S4PA on-line archive for inter-sensor value -added information G. Stephens, 2003 Giovanni A-Train data requested for retrieval

  19. A-Train Data Depot (ATDD) Steve Kempler, NASA/GSFC, GSFC Earth Sciences (GES) Data and Information Services Center (DISC) Description and Objectives ATDD provides homogeneous data access to heterogeneous data generated by the A-Train constellation sensor web. Objectives: • Gather heterogeneous sensor web measurements for intercomparison studies & future cross sensor communications. • Provide tools that allow first order temporal and spatial correlations of A-Train sensor data • Provide services (i.e., expertise) that will facilitate effortless access to and usage of ATDD data • Collaborate with scientists to facilitate the use of data from multiple sensors for long term atmospheric research Accomplishments - Sensor web implementation • Released Giovanni capability for Cloudsat/ MODIS intercomparisons, visualizations (vertical profiles aka “curtains”), and data access • Developed visualization for quick previews of atmospheric profiles data from MLS, MODIS, CloudSat, AIRS, and CALIOP. • Developed vertical regridding of CloudSat and CALIOP profile data from altitude to pressure grid. • Prototyped two-dimensional retrieval overplotting: Cloud top pressures: MODIS, AIRS; Vertical profiles: CloudSat & CALIOP • Published stories and image material on the web site. Sensor Web Work (beyond the scope of this project) - Cross sensor web communication • Content based searching to determine geophysical phenomena • Content based pointing, e.g., one senor node looks broadly and, on occasion, calls another sensor node for refined observations. • Real time cross sensor web communications Long term Schedule and Deliverables • Currently, beginning year 2 of a 2 year (possible 3 year) project; Level of effort ~1.5 FTEs per year • By July/07: Add Calipso, AIRS, OMI sensor node data • By October/07: Add HIRDLS, Parasol sensor node data • By January/08: Add CERES, TES sensor node data • By April/08: Add AMSU, AMSR-E sensor node data Above data displayed along Cloudsat or MLS track, as appropriate • By July/08: Capability to co-plot multi-sensor data on same axis • By January/09: All sensor node data co-registered

  20. Relative Positions of Afternoon Constellation Members In this scenario Aura and Aura have different WRS paths. CloudSat, CALIPSO, and PARASOL are on a path that is 215 km east of the Aqua path. The limb path for MLS (Aura instrument) is lined up with the Aqua ground track: 1:38 PM equator crossing time. Orbit Plane 197 km 197 km TES Limb track MLS Limb track Aqua MODIS 1150 km AIRS 825 km N Ground track (WRS Paths) CLOUDSAT CALIPSO 15 min (~ 52o lat) PARASOL 7 min (~ 26o lat) 373 km Aura 73 km Equator OMI 1300 km OMI swath is larger than MODIS or AIRS 394 km 172 km

  21. 6x7 km POLDER 1. 4 km Cloudsat 0.5 km MODIS Band 3-7 0.09 km CALIPSO Cloud 13.5 km AIRS IR; AMSU & HSB m wave OCO 1x1.5 km 5.3 x 8.5 km TES Washington DC USGS Map The Afternoon Constellation observational “footprints” vary greatly

  22. Backup:Selected AIST-05 Sensor Web Projects

  23. An Inter-operable Sensor Architecture to Facilitate Sensor Webs in Pursuit of GEOSS PI: Dan Mandl, GSFC Objective This project will develop the capability to generically discover and task sensors configured in a modular Sensor Web architecture, in space and in-situ, via the Internet. The proposed technology is thus well suited to assist future Earth science needs for integrating multiple observations without requiring the end-user to have intimate knowledge of the sensors being used. This project will demonstrate and validate a path for rapid, low cost sensor integration, which is not tied to a particular system, and thus able to absorb new assets in an easily evolvable coordinated manner. It will facilitate the United States contribution to the Global Earth Observation System of Systems by defining a common sensor interface protocol based upon emerging community standards. Vision for Space Sensor and Subsequent Science Data Access Via Generic Web Services to Form Sensor Web Approach Key Milestones • This project will help improve data acquisitions by • reducing response time and increasing data quantity and quality for the desired earth science data. This will be accomplished in the following ways: • Provide an interoperability standard • Enable instant discovery of available sensor resources • Enable the ability to direct other sensors • Enable the ability to specify how the available data should be delivered and combined • Development of relevant science & operations concepts and scenarios June 2007 • 1st demonstration EO-1 “discoverable”/taskable via Internet and the use of SensorML & EO-1 Autonomy SW Sept 2007 • Augment demonstration 1 with GMSEC framework in testbed for 2nd demonstration June 2008 • Integration of SensorML, IRC, GMSEC, cFE and CHIPS or testbed into 3rd demonstration Mar 2009 • Full capabilities demonstration, 4th demo Sept 2009 • Identification of Earth Science mission infusion targets Ongoing Co-I’s/Partners • Robert Sohlberg, Chris Justice, John Townshend /UMCP • Jeffrey Masek, Stuart Frye / GSFC • Stephen Ungar, Troy Ames / GSFC • Steve Chien / JPL TRLin = 3

  24. An Objectively Optimized Sensor Web PI: David Lary, UMBC/GEST Objective • Develop an autonomous Objectively Optimized Observation Direction System (OOODS) which will objectively optimize the observation schedules of a set of assets • Concentrate on the generic principles of how an OOODS would operate, its architecture, and development as a plug-in for a sensor web simulator/controller • Goal of the OOODS is to employ an objectively optimized data acquisition strategy for integrated observing systems that is responsive to environmental events for both application and scientific purposes Schematic of the Objectively Optimized Observation Direction System Approach Key Milestones • The system will be implemented by extending two existing technology components: • Analytical Graphics Incorporated (AGI) Satellite Tool Kit (STK), which will be used to develop the sensor web test bed • NASA award winning AutoChem system, which will be used for analysis • Sensor Web Data Interface and Preparation 12/2006 • Interim report for Simulator Component 1 12/2008 • Interim report for Simulator Component 2 12/2008 • Sensor Web Simulator Component 1 06/2009 • Sensor Web Simulator Component 2 06/2009 Co-Is/Partners • Michael Seablom / GSFC • Mark Schoeberl / GSFC • Stephen Talabac / GSFC • 04/19/07 TRLin = 4

  25. End-to-End Design and Objective Evaluation of Sensor Web Modeling and Data Assimilation System Architectures PI: Michael Seablom, GSFC Objective This project will: (i) design a sensor web architecture that couples current and future Earth observing systems with atmospheric, chemical, and oceanographic models and data assimilation systems; and (ii) build an end-to-end sensor web simulator (SWS) based upon the proposed architecture that would objectively assess the scientific value of a fully functional model-driven meteorological sensor web. The SWS will serve as a necessary trade studies tool to evaluate the impact of selecting different types and quantities of remote sensing and in situ sensors, to characterize alternative platform vantage points and measurement modes, and to explore rules of interaction between sensors and with weather forecast/data assimilation components to reduce model error growth and forecast uncertainty. Sample graphical user interface for the sensor web simulator. Approach Key Milestones • The proposed Sensor Web Simulator will be a large software system comprised of several large • Subsystems: • User Interface • Simulation Control • Simulation Analysis • Sensor Web Model • Simulated Observation Generator • Complete detailed design 02/2007 • Acquire GEOS5 & GSI codes from the GMAO 02/2007 • Complete re-engineering of OSSE 09/2007 • Command and Control / External Control • components preliminary design review 02/2008 • Design and implement software coupling from • Targeting component to External Control 09/2008 • Conduct OSSE for lidar instrument 09/2008 • Execute use case scenario with simulator 09/2009 Co-I’s/Partners • Stephen Talabac / GSFC • Brice Womack, Robert Burns / Northrop Grumman TASC • Joe Terry, Joseph Ardizzone / SAIC • Lars Peter Riishojgaard / UMBC Entry TRL= 2; Exit TRL = 4

  26. Satellite Sensornet Gateway (SSG) PI: Aaron Falk, USC Information Sciences Institute Objective • Enable rapid deployment of in-situ terrestrial sensors, particularly in remote or challenging environments • Provide a flexible, extensible interface between terrestrial in-situ sensornets and satellite communications networks. • Provide a structured and reusable management facility and suite of tools for remote sensor management. Distributed Sensor Network using Satellite Sensornet Gateway Approach Key Milestones • Design and prototype a open and scalable sensornet gateway that provides storage and aggregation of data from wireless sensors, reliable transmission to a central datastore, and sensor instrument management and control. • Blue-ribbon Science Advisory Board will shape system requirements based on research needs. • Design validation will be accomplished through a series of increasingly functional field deployments supporting Board members’ projects. • Assess candidate technologies 1/07 • Define architecture/functional allocation 3/07 • Bench test single-stream prototype 7/07 • Field test single-stream prototype 9/07 • Architecture revision 1/08 • Field test small-network prototype 6/08 • Realistic multi-network field deployment 6/09 TRLin = 3 Co-I’s/Partners None 4/07

  27. Optimized Autonomous Space - In-situ Sensorweb PI: WenZhan Song, Washington State University Objective Develop a prototype real-time Optimized Autonomous Space - In-situ Sensor-web, with a focus on volcano hazard mitigation and with the goals of: • Integrating complementary space and in-situ elements into an interactive, autonomous sensor-web. • Advancing sensor-web power and communication resource management technology. • Enabling scalability and seamless infusion of future space and in-situ assets into the sensor-web. Optimized Autonomous Space In-situ Sensorweb Approach Key Milestones •Develop a test-bed in-situ array with smart sensor nodes •Develop new self-organizing topology management and routing algorithms •Develop new bandwidth allocation algorithms in which sensor nodes autonomously determine packet priorities •Develop remote network management and reprogramming tools. •Integrate the space and in-situ control •Synthesize the sensor-web data ingestion and dissemination through the use of SensorML. •Demonstrate end-to-end system performance with the in-situ test-bed at Mount St. Helens, and EO-1 platform. • System requirements 3/2007 • System design 9/2007 • Testbed hardware assembly 3/2008 • System software design 3/2008 • Existing St. Helens Array Linked to EO-1 3/2008 • SensorML Development 9/2008 • Software implementation and testing 6/2009 • Field demonstration 12/2009 Co-Is/Partners • Frank Webb, Sharon Kedar, Steve Chien / JPL • Richard LaHusen / U.S. Geological Survey • Behrooz Shirazi / Washington State University TRLin = 2 TRLcurrent = 2 4/07

  28. QuakeSim: Enabling Model Interactions in Solid Earth Science Sensor Webs PI: Andrea Donnellan, JPL Objective Operational Concept • Improve the modeling environment for better earthquake forecasts, which will ultimately lead to mitigation of damage from this natural hazard. • Establish the necessary computational infrastructure • Develop optimal techniques for understanding the relationship between the observable space-time patterns of earthquakes and the underlying dynamics that are inaccessible or unobservable in nature. Includes data and model output Approach Key Milestones • GPS data federated into portal 8/07 • Parallel version of Virtual California (VC) 11/07simulation running on Columbia and Cosmos • Prototype InSAR database into portal 3/08 • Deployed on Cosmos and Columbia resources 10/08 • Fault database expanded to all of California 3/09 • Integrate Geographic Information System (GIS), Sensor Web, codes, and services 9/09 • Support for GIS and Sensor Web technologies 9/09 • Integrate real-time and archival sensor data with high-performance computing applications for data mining and assimilation • Federate sensor data sources, focusing on InSAR and GPS (Global Positioning System) • Extend QuakeSim to interact with high-end computing resources at Ames Research Center and JPL. Co-I’s/Partners John Rundle (UC, Davis) Geoffrey Fox (Indiana U.) Dennis McLeod (USC) Walter Brooks (ARC) Lisa Grant (UC, Irvine) Marlon Pierce (Indiana U.) Terry Tullis (Brown U.) TRLin = 3 4/07

  29. Land Information Sensor Web PI: Paul Houser, Institute of Global Environment and Society, Inc. Objective This project will develop a prototype Land Information Sensor Web (LISW) by integrating the Land Information System (LIS) in a sensor web framework. Through continuous automatic calibration techniques and data assimilation methods, LIS will enable on-the-fly sensor web reconfiguration to optimize the changing needs of science and solutions. This prototype will be based on a simulated interactive sensor web, which is then used to exercise and optimize the sensor web - modeling interfaces. In addition to providing critical information for sensor web design considerations, this prototype would establish legacy for operational sensor web integration with modeling systems. Enabling LIS to interact with sensor webs with open protocols and web Approach Key Milestones • This work will be performed in six steps: • Scenario development: a synthetic global land “truth” will be established • Sensor simulation: a model of a future land sensors will be established • Sensor web framework: sensor web communication, reconfiguration and optimization will be developed • Evaluation and optimization metrics: various land surface uncertainty, prediction and decision support metrics will be established • LISW experiments: to exercise and evaluate the system. • Sensor web design implications: design trade-offs • Scenario development March/2007 • Sensor simulation Sept/2007 • Sensor web framework March/2008 • Evaluation and optimization metrics Sept/2008 • LISW experiments March/2009 • Sensor web design implications August/2009 • Collaboration, Communication & Dissemination August/2009 Co-I’s/Partners • James Geiger / NASA-GSFC • Sujay Kumar, Yudong Tian / GEST-UMBC TRLin = 4

  30. Harnessing the Sensor Web through Model-based Observation PI: Robert Morris, ARC Objective The objective of this project is to build, integrate and demonstrate automated capabilities for model-based observing, a process of coordinating resources in a sensor web based on goals generated from Earth science investigations. Model-based observing will transform the sensor web into a cognitive web, a distributed, goal-directed sensing environment. The work will address three technical challenges: 1) transforming Earth science goals into plans for accomplishing those goals, 2) reconfiguring the web through the execution of the plans, and 3) generating new or revised goals from the results of previous observations. Architecture for proposed technology. Approach Key Milestones • Addressing technical challenges through the development of software capabilities for enabling three essential kinds of transformations • Extensive leveraging of the results of previous efforts • The extensive use of Earth science data to develop a robust demonstration platform Co-I’s/Partners • Jennifer Dungan / ARC • Petr Votava / ARC • Lina Khatib/ARC TRLin = 2

More Related