Monitoring and measuring the forest ecosystem is a complex challenge. Forest monitoring systems include measurement, reporting and verification (MRV) functions and aim to produce high-quality, reliable data on forests, including forest-carbon estimates that are critical to the battle against climate change caused by, among others, deforestation and degradation of the woods.  

Spatial data on deforestation and afforestation/reforestation is typically collected through satellite data, and changes can be monitored through a satellite land monitoring system. In addition, in many forest inventories, socio-economic information is collected to understand the anthropogenic impact on forests and their role in sustainable livelihoods. 

The University of Maine’s Wireless Sensor Networks (WiSe-Net) laboratory has developed a new method using AI and ML to make monitoring soil moisture more energy and cost-efficient. This could be used to make measuring more efficient across the broad forest ecosystem of Maine and beyond. The study was published on the 9th of August, 2022, in Springer’s International Journal of Wireless Information Networks.

Monitoring soil moisture 

Soil moisture is an essential variable in forested and agricultural ecosystems, especially considering the drought conditions in past summers. However, despite robust soil moisture monitoring networks and large, accessible available databases, the cost of commercial soil moisture sensors and the power they use to run can be prohibitive for researchers, foresters, farmers and others tracking the health of the land. 

The sensor developed by the researchers at the University of Maine uses AI to learn how to be more power efficient in monitoring soil moisture and processing the data. They had tied up with the University of New Hampshire and the University of Vermont researchers. The research was funded by a grant from the National Science Foundation. 

Understanding the sensor 

According to the principal investigator of the study, AI has the capacity to learn from the environment, predict the wireless link quality and incoming solar energy to efficiently use limited energy and make a robust, low-cost network run longer and more reliably.  

The model learns over time to make the best use of available network resources, which helps produce power-efficient systems at a lower cost for large-scale monitoring compared to the existing industry standards. In addition, the researchers ensured that the hardware and software of the models were informed by the science and tailored to the research needs. 

The moisture present in the soil is the primary factor driving tree growth. But rapid changes occur both daily as well as seasonally. Conventionally, expensive sensors are used to collect the data at fixed intervals. But these data might not always be reliable. This AI-powered model developed by the Maine researchers opens the door for a cheaper and more robust sensor with wireless capabilities, from which researchers and future practitioners can benefit.  

Although the system designed by the researchers primarily focuses on soil moisture, the same methodology could be extended to other types of sensors, like ambient temperature, snow depth and more, as well as scaling up the networks with more sensor nodes. 

The real-time monitoring of the variables requires different sampling rates and power levels. An AI agent can learn these and adjust the data collection and transmission frequency accordingly rather than sampling and sending every single data point, which is not as efficient.