Flood Early Detection System Using Internet of Things and Artificial Neural Networks

Natural disasters like floods are becoming more and more devastating every year due to increase in rainfalls and other factors induced by climate changes. The losses due to flood can be greatly minimized by the effective early detection systems. There are many traditional wireless sensor network methods currently available for this. But this paper gives a detailed study of how the current trending field of information technology called Internet of things is applied for an efficient implementation of the early warning flood detection systems. The paper describes how the flood can be predicted by extracting various parameters from the environment that contributes to the flood. A fully connected feed-forward artificial neural network is used here for the prediction purpose for giving early warning and communicating it to the target users. In the experiment, an Internet of things platform, hingspeak is used for real-time visualization of the sensor data. The alerts are sent to the preconfigured e-mail IDs and mobile numbers of the authorities and the communities without any delay.