An automated wireless sensor network based irrigation system

In irrigation systems watering large areas of crops is a difficult task as it often leads to wastage of water because of imprecise irrigation techniques. In order to overcome this problem, precision irrigation planning methods based on wireless sensor network technology help in automation of the irrigation process. In this research work, an automated irrigation system for an agricultural field with four different crops is proposed based on Artificial Neural Network (ANN), which will be used to train the irrigation system on the basis of properties such as air temperature, air humidity, soil temperature, soil moisture content, drought resistance, water stress level, rain level, flow of water etc. The water pump installed in the field can then decide when to be off/on and also determine the amount of water required by the crops using the ANN algorithm. A comparative analysis of the proposed work with another model which uses a fuzzy logic system is done using simulations in Matlab and it was observed that ANN based irrigation system can save 44.97 percent more water than Fuzzy Logic system when it comes to multi-cropping with timely response of the irrigation set up, with minimum delay.