Diary 43

Special Report on NSF ANI-9909218

30 May 2002

"Prototype Testing and Evaluation of Wireless Instrumentation for Ecological Research at Remote Field Locations by Wireless"

I

This 3 year project (1999-2002) has validated the great present, and future, value that emerging forms of FCC Unlicensed Spread Spectrum, UNII, and future 'smart radio' and ultrawide band protocol wireless technologies can be for the intermittent or continuous collection of biological and environmental data from sensors and data loggers emplaced in difficult and often seasonally inaccessible remote field locations. And the real time distribution of such data to researchers, globally, via the Internet. This technology can be in direct support of such proposed projects as NEON, with the ability to integrate real-time field data into displays and representation of already-processed and stored data.

Prior to the proof-of-concept experiments under this, and two prior, NSF Projects, researchers, such as the 1,200 scientists doing work in the 24 NSF funded Long Term Ecological Research (LTER) field stations, and those who have experiments in 150 Biological Field Stations under the cognizance Organization of Biological Field Stations were compelled to collect data manually from remote data loggers, or use costly, and limited capability and flexibility, FCC licensed radios.

II - Successful Areas of Experimentation

This project focused primarily on four widely dispersed LTER field locations representing a wide range of topographical and environmental field conditions.

* Bonanza Creek and Caribou Peak LTERS served by the University of Alaska, Fairbanks. Harsh climate, limited functionality of solar or wind power sources, seasonal inaccessibility of deployed data loggers. Labor and equipment costly (power boats, snow machines, ATVs) manual retrieval methods.

* Trout Lake Field Station, Wisconsin (North Temperate Lakes LTER). Study of fresh water regions. Dense woods, data sources out on lakes, winter inaccessibility, freezing and thawing of lakes, limiting data retrieval times.

* Luquillo Experimental Forest, LTER, Puerto Rico. Serving both University of Puerto Rico's El Verde and US Forest Service's Sabana Field Station in the rain forest. Rain forest vegetation, high canopy requiring large towers, deleterious effects of humidity and growths on technical equipment. Intense rainfall. Difficult accessibility of many sites. Many experiments difficult to monitor after being set up. And being even more difficult to change data collection parameters on data loggers from remote locations - such as in the face of a predicted event (weather, storms)

* Virginia Coastal Reserve LTER, Oyster, Virginia. Study on offshore uninhabited islands such as Hog Island, across 14 miles of water from the Field Station, variable and dangerous weather, tidal activity stranding researchers, inability to monitor by physical visits behavior of some animals, widely dispersed (10 km apart or more) sensor locations, inability to bring many types of students to the islands (liability and limited transport).

This project initially directly supported the science experiments being conducted at each field site, using a wide variety of off the shelf (but never designed for such scientific purposes, and very recent, Beta-Test) wireless devices, modified and interfaced, and powered in novel ways, to research equipment and attempted to bring the remote data signals back to the forward field stations.

Then, the researchers, grasping the potential of these new form of wireless requested over the last three years assistance in deploying experiments and data gathering tasks which could only be supported (season, location, scale) wirelessly.

High resolution video imagery, full motion video and sound, natural sound collection, interactive, remote data collection device management, simultaneous data-point collection, were enabled by advanced, low cost, high data rated devices that were hitherto practicably impossible. And the ability to 'see' underground and in depth behind surfaces are now being enabled - and requested by researchers.

Field data collection points were then interfaced to the Internet with wireless the 'last mile' so to speak, reaching remote researchers who are often able only to visit their experiments annually. As well as anyone else interested, including school children with net connections.

The advanced, digital, wireless technologies employed have only existed and been permitted by the FCC since 1985. They demonstrate great capabilities, but for field science, also great limitations - more attributable to FCC Technical Wireless Rules - originally designed for urban commercial, government or personal - not scientific - radio use, which rules to this date do no take into account the unique needs of field science. Part of the data gathering by this project also has been fed into the FCC so that they may revise their rules to better accommodate science in rural remote areas where wireless 'interference' is far less a problem than in urban areas. So this project has done a lot to foster change in US government rules for wireless support of science.

Radios with bandwidth as low as 19.2 to 20mbps have been deployed and tested. RF protocols from proprietary to 802.11b and a have been tested.

The pioneering aspects of this project have also been fed back into both Scientific Instrumentation and Radio Manufacturing companies, with recommendations, causing them to design and market radios more suited to the needs of field science and researchers than off the shelf earlier ones.

III - Major Finding

A major finding of this project has been that, with the proper design and networking of very small, low power, very low cost, singular-sensor 'routable signal' radios - ones which do not exist yet but which can be made - the entire widespread paradigm of the use of relatively few, costly, Data Loggers can be replaced by the deployment of thousands such miniature radios across research study 'fields,' with singular, interchangeable, sensors, very small processors, and with even remotely-rechargeable batteries (by laser light or microwave power). While data collection stations currently can cost (data logger, storage modules, sensors, radios, batteries, solar panels, apparatus) $5,000 each and up - such traditional 'centralized' data collection methods and economics, even wirelessly connected, severely limit their deployment 'at scale.' While the development of next generation unlicensed data radios from the findings of this study, which can cost as little as $100 or less together with one or two sensors, can permit wide scale deployment across water bodies or water-land margins, terrestrial fields, in tree stands - all capable of 'communicating' their data continuously or on demand. Producing scientifically more valid results. And still making that data instantly available world wide within whatever 'preliminary review' criteria researchers desire.

Example of the above are the problems of collecting light data on the floor of rain forests from many points simultaneously, the ability to monitor water levels in hundreds of 'pipes' in ground areas around lakes, the monitoring of multiple animal, bird, and insect sounds in field areas, the ability to monitor data every few feet up the trunks of large rain forest trees. The ability to deploy very small video capture devices in multiple locations.

A follow-on NSF grant is being proposed based on these finding. Which, if approved, can revolutionize field data collection coupled with global accessibility via the net.

A second finding is that 'power' for field equipment is a major, and still largely unsolved problem wherever the field locations for sensors, data collectors, and wireless are either in extreme cold climates (Alaska) or in heavily forested (rain forest or deep woods). When data loggers are manually visited frequently this problem is solved by carrying replacement batteries. But when the desirable goal of leaving large numbers of 'communicating sensors' in place for years, is envisioned, 'power' is an issue.

One observation was made during the project. One weakness in University and the Research Community preparation and support of field science has been revealed by this project. Neither PIs, or collaborators, or graduate students, or field station or university technical personnel are familiar enough with wireless, and in particular the capabilities that have emerged over the past 10 years, to use it to their advantage. In fact few technical staffs which run university networks know how to set up a wireless network, with the sometimes exception of internal wireless 'Lans,' which pose very different challenges from remote area field science supporting 'Wans'.

This project, through requiring field station technical staffs to accompany Wireless project personnel in doing 'site surveys,' experimenting, interfacing, and finally installing for long term use a variety of radios has amounted to a 'technology transfer' to many field science projects - whose personnel are themselves now extending wireless independently. One example is east of Nome, Alaska, where the University staff, having learned from us how to do it, has deployed - without our expert assistance - stations as far away as 80 miles through multiple relays and connected the data to the Internet.

Several workshops have been held, and now that the great value of wireless is understood, there is a substantial 'demand' to use this technology. But it is not except in the very simplest cases, plug and play. The last, oversubscribed, workshop drew attendees from French Polynesia, Panama, Mexico, Canada, Bermuda, as well as all over the United States.

There is a need for 'Wireless Technology Applications' programs (not to be confused with advanced RF engineering work) at every university with field science stations and experiments.

IV - Impacts on Society and Education

One, very recent, development has been to apply the findings of this study, in particular the potential for miniaturization and large scale deployment of wirelessly connected sensors to the threat of biological and chemical attack. The results of this study has been fed into the Multi-Sectored Crises Management Consortium which is under the umbrella of NCSA, and has been provided to other government agencies in early 2002. Collaboration with such agencies, and Russian scientists, as well as Industry, has occurred. The same wireless devices being designed for scientific data collection can be used for these national security purposes.

Education in Field Science at every academic level has been enhanced by the use of the wireless connectivity, interconnected to the Internet, using techniques developed and demonstrated through this project.

One example is the effort by the Virginia Coastal Reserve PI and assistants to help High Schools with which they are affiliated to learn about their science on Hog Island. The liability risks, and limited boat transportation precludes this LTER from taking all but a very few students out on that Atlantic island. Now, the PI goes out to Hog Island with a laptop, radio, video cam, and proper software and holds educational, two way, videoconferences with youngsters in schools. And can point his camera, microphone, up close to natural examples.

Another example is the ability, now of any school or college in the world with basic Internet access and web browsing software to monitor, track, record and use data from field sites in central Alaska.

The URL:

http://www.uaf.edu/water/projects/cpcrw/metdata/crrel/current.htm

gives periodic data from the CRREL Field and hourly updated video images of the area. All connected wirelessly through the University of Fairbanks (38 miles away) connection to the net. This required development of more advanced software techniques beyond the normal Data Logger specialized software used even over the Internet, to read data, and, or, upload changes.

Because of the low cost, while high bandwidth, internet-connectable wireless technology this project has pioneered along with appropriate interface and sensors links, schools, or even independent 'citizen scientists' can collect, and communicate data from their own research projects, fulfilling a prediction Dr. Larry Smarr made when he saw results of this wireless project at a 2001 NEON conference. He stated that schoolchildren and 'citizen scientists' could begin to 'collect and communicate field data' far beyond the ability of the limited number of doctoral level researchers or graduate students can presently. Likewise they can benefit in their own studies by having access to data collected elsewhere. Comparative, simultaneous, real time data from many places can be analyses from any connection to the Internet.

One other impact on society this project has had, is its pioneering of getting broadband Internet to remote and rural areas largely ignored by the wire-bound telephone companies. Where ever the staff of this project has set up wireless networks to support field science in remote areas, there has been a rapid 'migration' of the expertise displayed, into surrounding communities, which were convinced they could only be connected to the Internet at costly wired rates, or not at all.

The above phenomenon has also spread internationally, and many requests have been made for presentations, and support of foreign science projects. (based on the earlier 'Education by Wireless' NSF Funded project which led to this one, led to an NSF project in Ulaanbataar, Mongolia, that has now led to widespread use of wireless there, both for the support of US scientists there, and their society in general.

David R Hughes PI

Previous

Next