Electricity branches out

MIT researchers are working on a system that harvests energy from trees to power a network of sensors.


The US Forest Service currently predicts and tracks fires with a variety of tools, including remote automated weather stations. But these stations are expensive and sparsely distributed.


Additional sensors on the trees themselves could help to save more trees by providing better local climate data to be used in fire-prediction models and provide early alerts to fire hazards. However, manually recharging or replacing batteries at often very hard-to-reach locations makes this impractical and costly.


The new sensor system avoids this problem by using the trees as self-sustaining power supplies. Each sensor is equipped with an off-the-shelf battery that can be slowly recharged using electricity generated by the tree it’s on.


The system produces enough electricity to allow the temperature and humidity sensors to wirelessly transmit signals four times a day, or immediately if there’s a fire. Each signal hops from one sensor to another, until it reaches an existing weather station that beams the data by satellite to a forestry command centre in Boise, Idaho.


Scientists have long known that trees can produce extremely small amounts of electricity. But no one knew exactly how the energy was produced or how to take advantage of the power.


Now Shuguang Zhang, one of the researchers on the project and the associate director of MIT’s Center for Biomedical Engineering (CBE), and his MIT colleagues have figured that out: it’s all due to an imbalance in the pH between the tree and the soil it grows in.


Testing of the wireless sensor network, which is being developed by Voltree Power, is scheduled to begin in the spring on a 10-acre plot of land provided by the US Forest Service.


The researchers expect that they will need to instrument four trees per acre, noting that the system is designed for easy installation by unskilled workers. Right now, they are in the process of finalising exactly how the wireless sensor network will be configured to use the minimum amount of power.