To view this page ensure that Adobe Flash Player version 11.1.0 or greater is installed.

the storm‑water runoff in the wide expanse of hills and canyons surrounding Los Alamos in a cost-effective manner, and Frigo knew just how to do it. Frigo based her solution on radiofrequency transmission circuitry initially designed for networked Department of Defense surveillance satellites with many of the same design requirements as the storm-water samplers: transmission versatility to overcome line-of-sight issues, wide temperature range functionality, and extremely low power use. She set out to design and build field-deployable units that, in addition to being resilient against temperature swings and other weather extremes, could run on limited battery power for years on end and yet transmit more information farther than any comparable existing hardware. “The easiest thing would have been to use satellite modems,” says Frigo. “You measure something, send the information up to a commercial satellite on a subscription-based, pay-per-bit paradigm and then download it to your computer. We’ve had success with that approach in the past, but we couldn’t use it here. With all the data we would be sending from every sampling location, the satellite uplink would cost a fortune in hardware and data subscription fees.” Frigo and applied physics specialist Alexandra Saari opted for a multi‑pronged alternative. First, they would design each sampling unit to operate in a low-power idle mode except when transmitting. Second, they would give each unit a smart processor that could analyze the raw measurement data locally, filtering for important events and compressing the data to limit the duration of power-consuming transmissions to less than one second. But these advances alone would not be sufficient to overcome the power demands and prohibitive cost of satellite-based communications. Instead, the units were modified to transmit messages and alarms through local, low-power radio broadcasts. That meant each unit would be both sender and receiver, sampling station and relay station. In idle mode, the units would always be listening, allowing them to communicate with one another in a smart “mesh” network, intelligently taking turns passing data from one unit, or “node,” to another until the data reaches a common base station. They would also automatically reroute signals around busy or damaged nodes as changing field conditions may require. “There simply are no commercial off-the-shelf components that come close to meeting our needs in terms of cost, compactness, power use, data processing, transmission flexibility, and multi-hop mesh capability,” says Saari. “So we had to design our own system.” Indeed, the Los Alamos team now has more than 100 nodes operating in the surrounding canyon country, with each drawing power at a low enough rate that it can be accommodated by the sampler’s 12-volt battery, trickle-charged by a solar panel. Ultimately, the mesh will be expanded to 150 nodes, although it could accommodate another 100 more. “All we have to do to add a new node to the existing mesh is turn it on,” Saari says. “And if a node goes down for some reason”—she identifies lightning, raging floodwaters, and even elk antlers as potential culprits— “its neighbors will automatically seek a new route home around the missing node.” In addition, the nodes can be adapted to transmit and receive on different frequencies, should there ever be too much radio noise on a chosen frequency. This flexibility also allows them to operate within unusually narrow bandwidths, as required to obey regulations within controlled bandwidth environments, such as the region around Los Alamos National Laboratory. Farther afield, Frigo and others had already deployed an earlier version of the sensor-transmitter technology to support data collection in other diverse environments. The sensor units are helping ranchers in Northern New Mexico keep tabs on moisture, wind, and soil conditions across their sweeping ranchlands. They’re also recording and transmitting climate data from the northern reaches of Alaska and Sweden. It was the Los Alamos storm-water project, however, that advanced the system into a true self-forming, self-healing mesh network, allowing hundreds of nodes to coordinate reliably over mountainous terrain, from burning deserts to freezing icescapes. Perhaps that’s not so crazy for a system that had its origins in the ultra-remote, alternately searing and freezing wilderness of outer space.  —Craig Tyler Alexandra Saari configures an automated storm-water sampling station in the Jemez mountains near Los Alamos. 4 1663  October 2017