Come monsoon months and the news headlines are all about heavy rainfall and loss of hundreds of lives due to Flash Floods. In July 2019, at least 10 people were killed in Flash Floods when the river Brahmaputra burst its banks swamping more than 1,800 villages in Assam, India. In recent years, there have been reports of similar life-threatening incidents across the world.
Like its name signifies, Flash Floods happen in a flash, without prior warning. But is there still a way to predict Flash Floods and caution people who live in surrounding areas before the situation turns grave (often, a watery grave)? SignaGuard by MachineSense, an edge platform for sensor signal analysis, is working closely with the Information System Department of University of Maryland, Baltimore County, to develop an IoT system for advanced flood warning. Our beta testbed is Ellicot City, Baltimore, US, which was destroyed by the Flash Flood in May 2018.
First, let us understand how and why Flash Floods occur. This natural hazard takes place when the amount of rainfall is more than what the ground/soil can absorb or drain away. Roads turn into rivers and cars and people are swept away. Not all Flash Floods happen near rivers, many take place in urban areas where over-development, too much concrete and hard surfaces hamper drainage.
Using Motion and Vibration Diagnostics, SignaGuard studies the rate of rising water level from heavy rainfall, flow and turbulence in tributaries, geographical topography and rain forecasting. Water monitoring systems are deployed on bridges and lakes, for example, and the levels are monitored remotely, using ultrasonic and radar level IoT sensor technology.
By leveraging the technology of Internet of Things (IoT), data can be harnessed and analyzed real-time. When water levels rise above the danger mark, the police, paramilitary and disaster management agencies can be alerted.
This swift action can save hundreds of lives and damage worth millions. One small step for man, a giant leap for mankind.