Smart eco-friendly irrigation system based on advanced machine learning

Project aim: The project is aimed at developing a smart acgiculture irrigation system beased on advance mashine learning techniques.

 

About the project: In order to fulfil the world’s increasing demand for food, we must also take into consideration climate change and minimize the consumption of fuel, pesticides and other resources. Perhaps the most important natural resource is water. Today almost 80% of water is spent on irrigation and reduction of its consumption would have an extremely positive effect on the environment. Therefore, it is necessary to look for more effective, low-cost, and eco-friendly solutions. The aim of the project is to develop a smart integration system for irrigation based on advanced machine learning that would incorporate satellite images, soil maps, wireless field sensors and weather models. This model utilizes information about soil moisture (retrieved from sensors), predicted rainfall (from weather forecast models), soil’s ability to hold water (from soil analysis) and calculate the appropriate amount and time of irrigation on a daily bases. The effect of the whole process was closely monitored from satellite via spectral vegetation indices, which indicate the vigour and overall status of the plants, in order to ensure the maximum yield for the crop. Sensory readings were accompanied by satellite water-stress maps and localized weather forecast for the next 72 hours. The performance of pilot will be validated and assessed against specific indicators, such as water and fuel consumption per unit yield, economic efficiency of irrigation and yield quality indicator.