A smart hydroponics farming system using exact inference in Bayesian network

Abstract

Smart farming is seen to be the future of agriculture as it produces higher quality of crops by making farms more intelligent in sensing its controlling parameters. Analyzing massive amount of data can be done by accessing and connecting various devices with the help of Internet of Things (IoT). However, it is not enough to have an Internet support and self-updating readings from the sensors but also to have a self-sustainable agricultural production with the use of analytics for the data to be useful. This study developed a smart hydroponics system that is used in automating the growing process of the crops using exact inference in Bayesian Network (BN). Sensors and actuators are installed in order to monitor and control the physical events such as light intensity, pH, electrical conductivity, water temperature, and relative humidity. The sensor values gathered were used to build the Bayesian Network in order to infer the optimum value for each parameter. A web interface is developed wherein the user can monitor and control the farm remotely via the Internet. Results have shown that the fluctuations in terms of the sensor values were minimized in the automatic control using BN as compared to the manual control. The yielded crop on the automatic control was 66.67% higher than the manual control which implies that the use of exact inference in BN aids in producing high-quality crops. In the future, the system can use higher data analytics and longer data gathering to improve the accuracy of inference.