PLAN OF WORK

The Project is planned for 36 months.

 

Phase 1 will proceed from inventorying the current state of art sensors, wireless devices, satellite services, to making the initial simultaneous deployment of wireless links from sensors to the Trout Lake Field Station in Wisconsin and the El Verde Fiel Station in the El Yunque forest of Puerto Rico.

In each case careful technical site surveys will need to be taken to determine the location and characteristics of sensors, the pathways for digital wireless signals to travel, the data to be collected in terms of frequency, bandwidths required, and prot ocol formats, and the alternative sources of power sufficient to drive the sensors and the communications systems under the range of environmental conditions prevailing at the site, and over the time period desired by the researchers.

Then the appropriate commercially available equipment will be procured, the necessary original fabrication, interfacing, and coding will be accomplished in distant facilities, and bench testing will be accomplished. Then the systems will be installed on site, with personnel from the LTER team assisting, in order that the techniques of dealing with wireless can be learned by LTER technical personnel.

The tentatively planned - in collaboration with the LTER Site PI's - initial phase deployments include:

Luquillo LTER

1. Link three weather stations currently installed in the forest - El Verde and Bisley (each at 350 meters elevation) and Pico Del Este (1000 meters elevation). This will provide information on how environmental variables are affected by aspect (Bisley is windward, El Verde is leeward) and elevation (Pico Del Este is the summit of a high peak and other sites are nearer the base of the mountain.

2. Data from these weather sites will be linked wirelessly to the El Verde Field Site and through it (or in the case of Bisley through Sabena) to the Institute for Tropical Ecological Studies (ITES) and its web site in San Juan,

Trout Lake LTER

1. Link the battery powered, middle of Sparkling Lake tethered raft with underwater sensors, to the Trout Lake Field Station.

2. Test the feasibility of reaching 7 lake sensor sites from Utility towers on Musky Hill with relay to the Field Station. This will take much experimentation. 3. Connect the above sensory data through the network at Trout Lake Field Station, to the Internet via the 56kbps line now existing, or the future T-1 planned. Making the data available at least in the labs at the University at Madison, WI.

The First Phase is expected to take 6-8 calendar months with multiple trips to each site, Wisconsin and Puerto Rico after visits to selected sensor manufacturing companies surveying equipment and determining exact technical characteristics.

 

Second Phase

Observe the performance, robustness, reliability of the installed systems in Phase I, and troubleshoot, or redesign and replace.

Expand to the support of different types, from the above, sensors or data collection locations by additional and alternative types of radios, power, and communications techniques.

 

Luquillo LTER

This may include, in roughly the following priority expressed by the LTER team:

1. Monitoring streams continuously from current sensors which measure stream velocity, PH, conductivity, and ammonium, calcium, chloride, nitrate, potassium and sodium ions (via ion-selective electrodes). Three or more sites, possibly up to 12-15.

2. Collection of forest floor light data over time. Linking it to the web server for remote reading of data. This could include data on canopy distribution of light levels as well as on the forest floor. Develop methods for collecting data from multiple points in a measured land plot on the forest floor.

3. Added weather stations for greater coverage of Luquillo Mountains - which have the most variable weather in the world because of moisture laden trade winds.

4. Transmission of audio (microphone) collection of the sounds of a sub-species of Coqui frogs which inhabit El Toro Peak in El Yunque which are studied by night hike trips to the peak by collaborating LTER researchers with the USDA Forest Service.

Part of this phase will include interfacing such real-time audio data to web server data sections at the San Juan ITER web site location, for general use by other distant researchers. (These 'other researchers' may include, besides university-level resear cher, but also undergraduate science classes of K12 as well as general college students in both Puerto Rico and the United States which have web access to the net, and with the cooperation of the LTER projects, can attempt serious data collection and analysis)

 

TROUT LAKE LTER

This may include:

Fabrication of, or attachment to radio equipped buoys obtained by the research team that can stay in lake water year round and transmit from underwater sensors, such as water-depth pressure meters tethered to the bottom. Distribute sets of these and obser ve over time.

The North Temperate Lakes Site is planning to deploy several instrumented buoys at its primary study lake in the 1999-2000 time frame. We will link these instrumented buoys to the Field Stations and the Internet using wireless links. Funding by recent NSF grants to the LTER Project will determine how extensively (number of buoys or units can be deployed, where the buoys are purchased from LTER funds, while this wireless project funds will concentrate on providing the interfaces and linkages to the net) th ese can be accomplished.

1. Sensors for water temperature, dissolved oxygen, conductivity, PH, chlorophyll, ammonia, nitrate, chloride, and various weather variables.

2. Deploying, or modeling, deployment of a large number of inexpensive units to measure water level changes in lakes.

3. Carrying out similar wireless deployments, as in Phase I above for the Trout Lake area at the southern LTER water sites near Madison, Wisconsin. This will include training southern site technical personnel in how to deploy and interface radios f rom sensors to net access points. Phase II can be expected to take 6 -8 months with multiple trips to each site.

 

PHASE III

This is the 'push the envelope' phase, in which more complex and difficult tasks will be attempted.

The budget contains provisions for obtaining very innovative, small and mobile data collection, processing, transmission and data base presentation in the earliest stages of development by one of the most original technological inventors in the world, Ste ve Roberts of Nomad Research. We learned in the course of researching for this proposed project that he has early design plans for a series of highly integrated devices that can be carried on boats, both larger and small as canoes, capture and communicate environmental water data via multiple satellite and service channels. Support from this project will accelerate the creation of such original devices by July, 2000, insuring that they are made to specifications influenced by biological scientists who wou ld like this additional, mobile, water or overland data capture, auto-location sensing, and transmission capabilities ending in Linux based data bases without manual data entry. The design of the devices are not-proprietary, so the documentation will incl ude exacting plans for their construction out of readily available materials and operating systems at very low cost by researchers.

 

Some of the advanced techniques to be experimented with and models developed include:

1. Such as linking devices such as underwater cameras and video for observation of living creatures.

2. Audio capture and analysis of insects such as the Killer Bee infestation of El Yunque, whose sounds are distinct from other bee sounds.

3. Linking acoustic sensors capable of detecting zoo-plankton and fish sized particles.

4. Pursue deploying more advanced sensors that are not yet commercially available such as a CO1 Gas Flux system (a gas equilibrating chamber and a CO2 detector in association with an anemometer) to learn the differential between CO2 concentrations in surface water and the atmosphere above the lake as well as the wind speed one meter above the lake. Linking these back to the forward field site and the Internet.

5. Application to terrestrial based biological science projects, of devices originally fashioned for oceanographic research use. (there is, according to scientists canvassed for this project, not a lot of adaptation of oceanographic techniques and equip ment to terrestrial needs.)

6. From the communication standpoint, links to satellite-Internet services directly from sensor locations will be created, requiring the fabrication and integration of sensors, wireless, satellite ground station, power sources, interfacing and weather p roofing.

7. From the sensory data level, LTER PIs will identify data collection and analysis tasks that they have been unable to do with prior techniques. Appropriate sensors will be secured and interfaced, and wireless links from these will be established. Such advanced sensory work may include full-motion video capture, with wireless real time transmission, of wildlife or water creatures to a base system. Such as monitoring the nests and habitat of the endangered Puerto Rican Parrot. Or the use of sophisticated plankton measuring instruments and capturing and transmitting its data continuously. Or monitoring activity by underwater video cameras in stream entry points. Since radio waves do not propagate well through water, except at very low frequencies, there will be challenges in assembling useful systems which can communicate the real time data wirelessly. Dealing with forest canopies, reduced light levels, thick stands of trees, will require ingenuity and novel solutions to establish reliable, all weather c ommunications from sensors to remote networks and computer systems. One researcher in Puerto Rico would like to be able to observe, from home, 'local' conditions on El Yunque Peak at any time.

 

The connecting up of sensors, or sensors networked by terrestrial wireless to satellite communications base stations, where the base stations need to be small, consume little power whose battery sources are rechargable in the field, and then interface th em to the wireless network, and the sensors, and permit remote troubleshooting and diagnostics, will require substantial work and experimentation. Adapting very small low power consuming Linux systems operated in RAM, used by NASA on shuttle systems may b e promising ways to assemble routers for TCP/IP packets from multiple sensors.

If this project proceeds satisfactorily, centered originally on the two LTER sites in Wisconsin and Puerto Rico, applying the processes learned, and further developing wireless solutions to other needs can be considered. There are 19 other formal NSF LTR sites in the United States.

By this time in the project it can be expected, based on four years experience with the NSF Wireless Field Test Projects between 1995 and 1999, biological scientists and technical support personnel at the initial sites, and from other LTER Project sites, will come to understand the capabilities of wireless connectivity from sensory locations. And they will ask, or suggest other challenging experiments, which we can undertake so long as they add to the range of 'wireless models' we are attempting to develop.

 

Phase III can be expected to last 1 year at least, with multiple trips to each site, for both the new work and evaluating the results of earlier deployments of systems in the field.

Concurrent with all three phases described above, descriptions of the experiments and the developed system designs will be posted on the Project Web Site, as well as on individual LTER web sites currently in operation along with data links to the actual d ata-sensory devices, as approved by the LTER PIs, so that other researchers can both access the real and delayed time data, and have access to the designs of the systems used, described, with lists and general costs of materials, and what expertise is nee ded to duplicate the various models.

Also concurrent with the above, although probably not until Phase II, the PI will attend workshop and conference sessions of biological scientists and others, to report on the project, demonstrate the wireless access to sensors, and educate the field on i ts potential for their science. One such workshop is being proposed by LTER personnel on spread spectrum technology at an LTER 'all-scientists' meeting for August, 2000. 800 are expected.

This will be supplemented by written articles, and a video tape for distribution on request, of the project.

Other Work

One important task will be the 'technology transfer' from the PI and the Project to both the PI, Co-PIs assigned to the biological sciences projects by the LTER PIs, on the details of actual design, fabrication, wireless link testing, and deployment of th e systems. So that, when the Project is over, LTER and other biological science staffs can duplicate the models for their own science and sites.

As an extension of the value of the university basic biological research being done and the wireless links of real data to the Internet, selected K-12 schools, at the science level in middle and high school, will be invited to use the Internet to link to selected study projects on the LTER Web sites, and to a greater or lesser degree 'participate' in the analysis of the gathered data. This could lead to the participation of qualified advanced students and perhaps school science teachers in some of the LTE R projects in support of the PIs.

Graduate students from the biological sciences work in the LTERs can be expected to be offered by the PIs to work with the wireless technologies, particularly when they are available on site during summer periods.

Throughout the project, the PI, after using and adapting industry sensors and data logging instruments, will, as was done for Pasco Scientific during and after a seminal 'field science by wireless' experiment was conducted at the school level in Montana ( NSF Wireless Field Tests for Education), communicate critiques of their instruments used in this unique way back to industry. So that they may better design equipment used by scientists so they may be linked, with long period deployments, to wireless networks.