Gordon Blair, a professor of computing at Lancaster University, has a keen interest in how distributed data-collection systems can be used to better understand ecosystem functions, and to mitigate environmental harm. So, when the Engineering and Physical Sciences Research Council (EPSRC)—a science and technology research arm of the United Kingdom’s government, similar to the National Science Foundation in the United States—put out a call for proposals for researching the applicability of distributed, sensor-based communication systems within the natural (as opposed to urban) environment, it was music to Blair’s ears.
“I’m very interested in environmental challenges,” Blair explains, “and my skills are in distributed systems technology, which can help environmental scientists make sense of environmental systems.” In the EPSRC’s request for proposal, he saw “an opportunity to work on [the] IoT in the natural environment—because everyone else is working on [IoT solutions in] smart cities.”
EPSRC awarded Blair a $265,000 grant to deploy an IoT system in a rural environment, in order to assess its efficacy in monitoring the health of livestock, as well as that of the landscape and riparian environment. Blair then assembled a team of data and environmental scientists from Lancaster University, the United Kingdom’s Centre for Ecology and Hydrology, the British Geological Survey and Bangor University. The 18-month project began in December 2014.
A number of news reports have focused on one common element of the landscape in Conwy, Wales, where the experiment will take place: sheep. Numerous headlines have reported that the project’s goal is to turn the animals into Wi-Fi hotspots by outfitting them with “digital collars.” But that is hardly the point of the project, Blair explains. For one thing, he says, sheep would make lousy hotspots, since they move while grazing and tend to cluster in groups. Beyond that, the aim of the project is not just to establish sensor and communication networks within a rural setting, but also to use those technologies to better understand that setting and its biological systems—including sheep.
“What we’re trying to do is measure a range of things, and the sheep are just one of those things,” Blair says. “We’re looking at water systems, soil, animals, pollution and how they all interact with each other.”
These factors are important to the health of the landscape, the livestock and the humans who rely on that particular watershed. Conwy, where the project is being deployed, is within the watershed of Mount Snowdon, the highest peak in Wales. The landscape between Snowdon’s summit and the Irish Sea harbors 17 different vegetation zones, with a range of rich soils types, all within an area of 30 to 40 miles. “It’s a very rich range of ecosystems,” Blair states.
As for how they will monitor environmental variables within those landscapes, Blaire and his colleagues are still in the planning stage, developing “storyboards” to describe the software architecture that will support data collection from a range of wireless sensors measuring such factors as soil moisture, soil composition (by measuring nitrogen and phosphorous levels), location (via GPS) and movement (using accelerometers and tilt sensors). The researchers will assemble and distribute the sensors—which will most likely communicate via a mesh network using the ZigBee protocol—themselves, utilizing Arduino boards and electronics. Blair expects the team to deploy between 50 and 100 sensors.
Because residents rely on water from rivers that move through agricultural areas, Wales has been working with farmers to ensure that runoff from livestock waste and pesticides sprayed on crops does not enter the water supply. “The U.K. government is paying farmers to add transition zones between where they keep livestock and rivers,” Blair explains, “by doing things such as planting trees [along the river].”
Pathogens from feces can enter waterways in places where there is no transition zone, or where there is insufficient soil to capture and decompose the animal waste. The amount of moisture soil holds will impact whether rainfall is absorbed or if it will run off into the nearest stream or river, bringing pathogens from feces in fields along with it. By using IoT technology, Blair hopes that farmers will be able to better understand and model the impacts of rainfall on the water quality within their watershed.
Sensors placed along riverbanks could monitor flow levels—which, along with data collected from a weather station already in place on the farm where the pilot is taking place, could help the researchers better understand how and when pollutants enter the waterways. The sensors could also help them mitigate agricultural-related environmental damage, by indicating excessive levels of nitrogen or phosphorous in the soil.
As for those sheep, Blair says that by attaching tilt sensors to the animals, farmers may be able to monitor for signs of illness, such as foot rot, a contagious disease that generally causes inflicted animals to limp, and to kneel down while grazing. Another illness causes the sheep to walk in circles, which sensors could detect as well.
But how will the researchers transmit all of this data from the farm to the lab? That is still to be determined, Blair says. “We’re engaging with some telecom providers who are interested in [expanding into] rural coverage,” he reports. “There is a huge industry growing around smart agriculture, and some wireless telcos see the potential in that space. So hopefully, we can work with them. If not, we’ll need to [use] satellite communications, or find a farmer with broadband who is willing to share.”