Flooded streets — A crowdsourced sensing system for disaster response: A case study

Abstract

This study is about the disastrous flooding of an Indian metropolitan area of Chennai when the rain had nearly broken the record of 100-years with 374 mm rain falling on December 1, 2015, virtually breaking the November monthly average of 407.4 mm in a day. This city with a population of approximately 6.7 million people came to a standstill. Astonishingly, one of the biggest software development hubs in India was struggling for data and tools to identify which parts of the city were most affected and vulnerable to such climate phenomena. A group of software engineers quickly came up with an idea of using a Flood Map tool managed through crowdsourcing to help the citizens of Chennai and prevent further casualties. They developed a map-based tool called Flooded Streets to report flooded streets using OpenStreetMap (OSM) data. Using this Flood Map tool, anyone in the crowd could click on a street if they knew it was flooded and update the map information. Within the following 24 hours, over 2,500 streets had been reported as flooded by the citizens of Chennai using the Flood Map tool. An ordinary citizen could zoom into a locality, visualize which streets are reported as flooded and decide their next course of action. This map was also a great aid tool for relief and aid workers to track the flooded paths and provide appropriate aids in that area. The map consists of a base layer of low-lying areas created using elevation models from ISRO and NASA, and flooded areas from UNITAR. The map interactivity was built using Mapbox GL and hosted on GitHub. This crowdsourced sensing system is an extraordinary example of disaster response using the crowdsourcing concept, which potentially helped millions of people with the minimum time and resources but with great crowd contributions from both experts and non-experts.