Towards a Smart Low-cost Deluge Forecasting System

Over the years, floods caused by heavy rainfall have resulted in massive destruction in the world, leading to the tragic loss of many lives alongside severe economical setbacks. Studies have shown that early prediction of such scenarios would mitigate these effects. The prediction can be done using different machine learning algorithms that analyze the collected data from local and global databases. This project proposes a low-cost IoT based flood monitoring and alerting system that collects data using several sensors and a microcontroller, then uses a machine learning model previously trained on remote sensing data of that specific area to predict the amount of precipitation based on the real-time collected data. The processed data is presented through a web page and the raw data can be used by the authorities to take protective measures. This project can help organizations collect high-accuracy data with low costs, while individuals can also use the data in their research. The proposed system will also open the door to developing such setups for different types of data, which would drive more innovation in the field of studying the Earth's atmosphere and its associated phenomena using machine learning and IoT techniques.