Real-Time Soil and Crop Sensors - How Well Do They Work? 961756
Development of sensors, particularly sensors specifically developed for Site-Specific Crop Management, currently lags the other enabling technologies. Real-time sensors, i.e., those that can sense a soil or plant parameter and produce an estimate of the value of the parameter within a few seconds, are under development.
Our research has exploited two technologies - spectral reflectance and ion-selective membranes, for sensing of soil parameters. We found that optical estimation of soil organic matter (SOM) and soil moisture content can be accomplished with wide-band and narrow band spectral reflectance data. A rugged portable NIR sensor was developed which was able to predict SOM (r2 = 0.89, SEP = 0.40% SOM) and soil moisture (r2 = 0.94, SEP = 1.88%). A system using ion-selective membranes, field effect transistors, and flow injection analysis was able to predict soil nitrates in manual soil extracts with r2 > 0.90.
Research on the sensors for crop parameters has included the use of a light source and photodiodes to measure grain flow rate on a combine. A spring-loaded rod attached to a rotary potientiometer, was shown to count corn plants at harvest with errors in the 5% to 10% range. Preliminary results with a photoelectric sensor appears to have comparable accuracies, while also providing a measure of plant stem diameter.