IN23A-1067
SensorKit: An End-to-End Solution for Environmental Sensor Networking
Modern day sensor network technology has shown great promise to transform environmental data collection. However, despite the promise, these systems have remained the purview of the engineers and computer scientists who design them rather than a useful tool for the environmental scientists who need them. SensorKit is conceived of as a way to make wireless sensor networks accessible to The People: it is an advanced, powerful tool for sensor data collection that does not require advanced technological know-how. We are aiming to make wireless sensor networks for environmental science as simple as setting up a standard home computer network by providing simple, tested configurations of commercially-available hardware, free and easy-to-use software, and step-by-step tutorials. We designed and built SensorKit using a simplicity-through-sophistication approach, supplying users a powerful sensor to database end-to-end system with a simple and intuitive user interface. Our objective in building SensorKit was to make the prospect of using environmental sensor networks as simple as possible. We built SensorKit from off the shelf hardware components, using the Compact RIO platform from National Instruments for data acquisition due to its modular architecture and flexibility to support a large number of sensor types. In SensorKit, we support various types of analog, digital and networked sensors. Our modular software architecture allows us to abstract sensor details and provide users a common way to acquire data and to command different types of sensors. SensorKit is built on top of the Sensor Processing and Acquisition Network (SPAN), a modular framework for acquiring data in the field, moving it reliably to the scientist institution, and storing it in an easily-accessible database. SPAN allows real-time access to the data in the field by providing various options for long haul communication, such as cellular and satellite links. Our system also features reliable data storage and transmission, using a custody transfer mechanism that ensures data is retained until successful delivery to the scientist can be confirmed. The ability for the scientist to communicate in real-time with the sensor network in the field enables remote sensor reconfiguration and system health and status monitoring. We use a spiral approach of design, test, deploy and revise, and, by going to the field frequently and getting feedback from field scientists, we have been able to include additional functionality that is useful to the scientist while ensuring SensorKit remains intuitive to operate. Users can configure, control, and monitor SensorKit using a number of tools we have developed. An intuitive user interface running on a desktop or laptop allows scientists to setup the system, add and configure sensors, and specify when and how the data will be collected. We also have a mobile version of our interface that runs on a PDA and lets scientists calibrate sensors and "tune" the system while in the field, allowing for data validation before leaving the field and returning to the research lab. SensorKit also features SensorBase, an intuitive user interface built on top of a standard SQL database, which allows scientists to store and share their data with other researchers. SensorKit has been used for diverse scientific applications and deployed throughout the world: from studying mercury cycling in rice paddies in China, to ecological research in the neotropical rainforests of Costa Rica, to monitoring the contamination of salt lakes in Argentina.
IN23A-1068
Distributed Computing and MEMS Accelerometers: The Quake Catcher Network
Recent advances in distributed computing provide exciting opportunities for seismic data collection. We are in
the early stages of implementing a high density, low cost strong-motion network for rapid response and early
warning by placing accelerometers in schools, homes, offices, government buildings, fire houses and more.
The Quake Catcher Network (QCN) employs existing networked laptops and desktops to form a dense,
distributed computing seismic network. Costs for this network are minimal because the QCN uses 1) strong
motion sensors (accelerometers) already internal to many laptops and 2) low-cost universal serial bus (USB)
accelerometers for use with desktops. The Berkeley Open Infrastructure for Network Computing (BOINC!)
provides a free, proven paradigm for involving the public in large-scale computational research projects.
The QCN leverages public participation to fully implement the seismic network. As such engaging the public
to participate in seismic data collection is not only an integral part of the project, but an added value to the
QCN. The software provides the client-user with a screen-saver displaying seismic data recorded on their
laptop or recently detected earthquakes. Furthermore, this project installs sensors in K-12 classrooms as an
educational tool for teaching science. Through a variety of interactive experiments students can learn about
earthquakes and the hazards earthquakes pose.
In the first six months of limited release of the QCN software, we successfully received triggers and
waveforms from laptops near the M 4.7 April 25, 2008 earthquake in Reno, Nevada and the M 5.4 July 29,
2008 earthquake in Chino, California (as well as a few 3.6 and higher events). This fall we continued to
expand the network further by installing seismometers in K-12 schools, museums, and government buildings
in the greater Los Angeles basin and the San Francisco Bay Area. By summer 2009 we expect to have 1000
USB sensors deployed in California, in addition to any current or new laptop users.
http://qcn.stanford.edu
IN23A-1069
Monitor System for Space Electromagnetic Environments: Sensor Network in Space
We propose a monitoring system for space electromagnetic environments. We address it MSEE(Monitor System for space Electromagnetic Environments). The MSEE is a kind of the sensor network system in space. It consists of palm-sized sensor nodes, which are randomly distributed in the target area. The sensor node carries a compact plasma wave receiver as well as other necessary components such as communications and digital processing units. The observed data are transferred to the center station such as space stations or satellites/rockets through the ad-hoc network system. The objective of the MSEE is to observe plasma wave activities in multiple-points. Since space plasmas are essentially collisionless, kinetic energies of plasmas are exchanged through plasma waves. This means the plasma wave activities well reflect the variation of the environments in space which is filled with space plasmas. The targets of the MSEE are the artificial disturbances due to human activities in space as well as natural plasma waves. The MSEE provides us with the information on the three dimensional variation of the space electromagnetic environment in the target area. Recently, we have developed the prototype of the sensor node. In the prototype sensor node, small electric and magnetic field sensors with enough sensitivity and their small preamplifiers are installed. We also develop the small plasma wave receiver using the analogue ASIC technology. The necessary analogue components of plasma wave receivers are realized in one-chip ASIC with the size of 3mm x 3mm. The system of the sensor node is controlled by the one-chip computer. Under its control, communications and location identification are done using the wireless network technology. In the present paper, we introduce the MSEE system and its design. We also report the current status in our developing the small size plasma wave receivers with their sensors and the technique of the location identification of each sensor node.
IN23A-1070
Experiences with engineering, making and deploying sensor networks
Engineers and computer scientists will usually persuade themselves that producing a sensor network is matter of design, test and deploy. After several deployments in and on Glaciers within the Glacsweb project we are in a better position to understand the reality of producing sensor networks for real-world deployments. Not only does the electronics design, programming, management and logistics have to be perfected but a full understanding of the geoscience user's priorities and needs have to be an integral part of the system. This talk will outline the achievements of the 2008 Iceland subglacial probe deployment concentrating on the unexpected things which can affect the success of such a system. This includes the design of a new sensor node which is designed for low power, easy programming and high flexibility.
IN23A-1071
Toward a Cyberinfrastructure for the Ocean Observatories Initiative: Enabling Interactive Observation Within the Oceans
The Ocean Observatories Initiative (OOI) is an environmental observatory covering a diversity of oceanic environments, ranging from the coastal to the deep ocean. It is currently in the final design phase, with construction planned to begin in mid-2010 and deployment phased over five years. The key integrating element of the OOI is a comprehensive cyberinfrastructure whose design is based on loosely coupled distributed services, and whose elements are expected to reside throughout the OOI observatories, from seafloor instruments to deep sea moorings to shore facilities to computing and archiving infrastructure. There are six main components to the design comprising the core capability container, consisting of four elements providing services for users and distributed resources and two infrastructural elements providing core services to them. The sensing and acquisition component provides capabilities to acquire data from and manage distributed seafloor instrument resources, including their interactions with each other and with the infrastructure power, communication and time distribution networks. It includes services to publish instrument data and a repository for instrument behaviors and processes. The data management component provides capabilities to distribute and archive OOI data, including cataloging, versioning, metadata management, and attribution and association services. A core component will be an OOI-standard data/metadata model. The analysis and synthesis element provides a wide range of services to users, including control and archival of models, event detection services, quality control services, and collaboration capabilities to enable the creation of virtual laboratories and classrooms. The planning and prosecution element gives the ability to plan, simulate and execute observation missions using taskable instruments, and is the cyberinfrastructure component that turns the OOI into an interactive observatory. The remaining elements are the common operating infrastructure (COI) and the common execution infrastructure (CEI). The COI provides core services to manage distributed, shared resources in a policy based framework, including a distributed service infrastructure for the secure, scalable and fault tolerant operation and federation of the operational domains of authority comprising the OOI. It includes capabilities to manage identity and policy, manage the resource life cycle, and catalog/repository services for observatory resources. It also manages interactions with resources on an end-to-end basis. The CEI provides an elastic computing framework to initiate, manage and store processes that may range from initial operations on data at a shore station to the execution of a complex numerical model on the national computing infrastructure.
IN23A-1072
Life Under Your Feet: A Wireless Soil Ecology Sensor Network
Wireless sensor networks have the potential to revolutionize soil ecology by providing abundant data gathered at temporal and spatial granularities previously impossible. We will present the design and the results we gained by deploying multiple experimental networks for soil monitoring over the past three years. Our current deployments include turtle nest monitoring, soil temperature and moisture monitoring related to an urban CO2 flux tower and soil respiration measurements. The networks are now part of an end-to-end system.
IN23A-1073
Real World Example of a Hazard Warning System based on a Network of Autonomous High Rate, Low Latency GPS Sensors
The California Real Time Network consists of 80 autonomous geodetic-quality continuous GPS stations
distributed at nominal 20 km spacing throughout Southern California, with plans to expand statewide with a
spacing of 80 km. The GPS receivers collect data at a sampling rate of 1 Hz, with capability up to 20 Hz. The
data are streamed to a server at Scripps over dedicated spread spectrum radios and existing microwave
communication links (such as provided by UCSD's HPWREN project), with a latency of less than 1 s. We have
also experimented with streaming data over cellular modems using commercial service providers. CRTN
provides an operational test bed for an early warning system (EWS) for geological (earthquake, tsunami,
volcano, landslide) and meteorological (flood) hazards. At the server end, we use a GridSphere and
JavaServer Pages-based web portal environment with components developed under several NASA-funded
projects by Scripps and JPL, enabling users to select, view, manipulate and download GPS data products. An
earthquake EWS makes use of and extends some of the components of this web portal. CRTN has
constructed portlets to display: 1) strain calculated from the changes in displacements of the GPS stations
within a Delaunay triangulation mesh; 2) archived animations of strain maps of seismic events, both actual
and simulated; 3) time series of the displacements at a given site; and 4) rapid earthquake models including
moment magnitude and location. Another component of the EWS will generate messages to notify registered
First Responders upon the detection of seismic events using a lightweight component of the Geophysical
Resource Web Services (GRWS) framework developed as part of several NASA projects. Strain and
displacement data will be available from within the portal so that users may import data into the portal's web
service-based configurable filters and models. We discuss the merits of various components of our sensor
network and issues that have arisen in building and deploying the software applications.
http://geodemo-
c.ucsd.edu/gridsphere/gridsphere
IN23A-1074
A Model-Driven Sensor Web Simulator
Although numerous strides in weather forecasting were achieved in the past forty years, many of the operational concepts for today's observing systems remain essentially unchanged. Improvements have been suggested to address this, particularly the use of a model-driven sensor web architecture. Such a sensor web would potentially consist of ground systems, weather models, sensors and other components. The sensor web would enable cooperative, real-time measurements and targeted observations that can be used to improve operational efficiency and the forecast model's predictive skill. Implementing such a sensor web system, however, would impose significant costs and risks. Therefore, NASA Goddard is developing a tool called the Sensor Web Simulator. This tool is designed to simulate the behavior of a real-world, model-driven sensor web. The simulator leverages technology from Observing System Simulation Experiments (OSSEs) to create simulated observations for the desired observing platforms. Through the use of a workflow tool and data services, the Sensor Web Simulator integrates multiple models and software components and allows users to configure options and then execute a sensor web scenario. Although the work is in its early stages, the potential for its benefits have already been demonstrated. An initial experiment using sensor web concepts shows how a model-driven operations concept with the GWOS lidar could minimize the required number of lidar shots without compromising the information of the model's atmospheric state. Future work on the Sensor Web Simulator was proposed to expand the system beyond a "breadboard" system of components into an integrated mission design tool for scientists and engineers.