Artificial intelligence used to better monitor Maine’s forests
Monitoring and measuring forest ecosystems are complex challenges as software, collection systems and computing environments require increasing amounts of energy. Now the University of MaineThe Wireless Sensor Networks Lab, or WiSe-Net, has developed a new method of using artificial intelligence and machine learning to monitor soil moisture with less energy and cost. The method could be used to increase the effectiveness of measurements in forest ecosystems in Maine and beyond.
Soil moisture is an important variable in forest and agricultural ecosystems, especially in the recent drought conditions of Maine summers. Despite strong soil moisture monitoring networks and large freely accessible databases, the cost of commercial soil moisture sensors and the power they consume can be prohibitively expensive for researchers, foresters, farmers and others who monitor the health of the earth.
WiSe-Net researchers have designed a wireless sensor network that uses artificial intelligence to learn how to be more energy efficient in soil moisture monitoring and data processing. The work was funded by a grant from the US National Science Foundationit is EPSCoR program, designed to promote scientific progress nationwide.
AI can “use limited energy efficiently and keep a low-cost, robust network running longer and more reliably,” says Ali Abedi, an electrical and computer engineer at the University of Maine. The software learns over time how to make the most of available network resources to produce energy-efficient systems at lower cost for large-scale monitoring.
WiSe-Net has also collaborated with Aaron Weiskittel, Director of the Center for Sustainable Forest Research, to ensure that hardware and software research is informed by science and relevant to research needs.
“Soil moisture is the primary driver of tree growth, but it changes rapidly, both daily and seasonally,” Weiskittel says. “We didn’t have the ability to monitor this effectively at scale. A cheaper, more robust sensor with wireless capabilities opens the door to future applications.”
The study is published in the International Journal of Wireless Information Networks.
Although the system focuses on soil moisture, the same methodology could be extended to other types of sensors for measurements of ambient temperature, snow depth and other variables. More sensor nodes could scale networks.