Wireless Sensor Networks and Actionable Modeling for Intelligent Irrigation

Wireless Sensor Networks (WSNs) are emerging as an invaluable tool in agriculture. Current WSNs are good at reporting sensor data and many provide simple tools for viewing the data. The current approach assumes that growers are familiar with how the data should look and they are trained to use the data. This work extends the usage of WSNs to provide growers with actionable results based on the data. We are currently implementing plant physiological models for specialty crops into the CMU SensorWeb system. These models predict water usage, allowing growers to precisely control irrigation schedules. The sensor nodes used in this work also have onboard control functionality for automating crop irrigation based on the model’s output. In this paper we present a framework for integrating models into WSNs and initial field results from this model driven approach.