Sensor networks have been applied to various research
areas at a number of academic institutions. In particular, environmental monitoring
has received a lot of attention with major projects at UCB, UCLA and other places.
In addition, commercial pilot projects are staring to emerge as well. There are
a number of start-up companies active in this space and they are providing mote
hardware as well as application software and back-end infrastructure solutions.
The University of California at Berkeley in conjunction with the local Intel Lab
is conducting an environmental monitoring project using mote based sensor networks
on Great Duck Island off the coast of Maine. This endeavor includes the deployment
of tens of motes and several gateways in a fairly harsh outdoor environment. The
motes are equipped with a variety of environmental sensors (temperature, humidity,
light, atmospheric pressure, motion, etc.). They form a self-organizing multi-hop
sensor net work that is linked via gateways to a base station on the island. There,
the data is collected and transmitted via a satellite link to the Internet. This
setup enabled researchers to continuously monitor an endangered bird species on
the island without constant perturbation of their habitat.
The motes gather detailed
data on the bird population and their environment around the clock which would.The
Intel Mote has been designed after a careful study of the application space for
sensor networks. We have interviewed a number of researchers in this space and
collected their feedback on desired im-provements over currently available mote
designs. A list of requests that have been repeatedly mentioned includes the following
key items: o Increased CPU processing power. In particular, for applications such
as acoustic sensing and localization additional computational resources are required.
o Increased main memory size. Similar to the item above, sensor network applications
are beginning to stretch the limits of existing hardware designs. This need is
amplified by the desire to perform localized computation on the motes.