Smart Agriculture System of Flood Monitoring and Mitigation Using Live Data for Flood-Prone Area

Abstract

This research proposes a solution to improve the system for monitoring relevant environmental parameters using sensors for flood mitigation. Sensors are used to collect data regarding farm flood situation. The collected data are trained for a classification model to activate the solar-powered water pump to mitigate flood incidents in a flood-prone area. The system helps farmers to monitor real-time environmental parameters relevant to farming operations and flood including soil moisture level, water level, and water flow speed in a nearby canal that provides water to the farm. To reduce flood damage, the system assists to drain the excessive water to prevent prolonged submerging of the crop. The devices are designed to use the electricity from solar power, so the system is practically used outdoor where an electricity cord is difficult to setup. Experimental results show that the sensing data from the deployed sensors are accurate. The generated prediction models give the high performance with average of 1.0, 0.97, and 0.93 F-1 score for no-flooding, mild-flooding, and severe flooding respectively.