Computation-Time and Mean-Squared-Error Trade-off of Calibration Curves for Electrical Soil Conductivity Smart Sensor

Precision Agriculture and Internet of Things are integrating more and more their solutions. In this context, it is very important to evaluate the IoT solutions of demanding low-cost and low-power consumption capable to be applied in Precision Agriculture. An embedded processor with low cost and the most energy-efficient is the ARM Cortex-M0+ processor that has been widely used for building up IoT platforms. For its turn, measuring soil electrical conductivity is one of the most measured variable in Precision Agriculture. In this work, it is described a developed soil electrical conductivity smart sensor based on a low-cost and commercial soil conductivity sensor and on an ARM Cortex-M0+ processor. Calibration curves of the proposed soil conductivity smart sensor have been obtained and a performance analysis considering the computation time and the Mean Squared Error for the curves has been conducted. As a result, a particular time computation method for ARM Cortex-M0+ is described showing a high resolution and high precision and a trade-off between the proposed Smart Sensor's Computation-Time and Mean-Squared-Error could be obtained.