Bangladesh | Crowdsourcing flash-flood warnings that race the waters downstream
21 Aug 2014 by Anu John
Innovation is about taking risks, refusing to bow down to the impossible. Right?
Well, we took the risk and failed at get-go.
“This idea is not going to work. Please don’t waste the money.” That’s how one government minister dismissed our idea outright after we approached him for support.
We wanted to use mobile phones – now ubiquitous in Bangladesh – to crowdsource flash-flood warnings. Alerting people about the imminent onrush of flash floods in northern Bangladesh (located along the highest rainfall zone in South Asia) would help save lives especially those of the elderly and children from the rush of oncoming waters.
Bangladesh’s early warning systems to predict more conventional floods (ones that involve gradually rising water levels in the country’s major rivers) have improved by leaps and bounds in the past decades. But lack of cross-border cooperation in hydro-meteorological data means fast-moving waters that descend from the Himalayan foothills in India can sweep away houses, flatten crops and leave a trail of death and destruction in a matter of hours.
That’s what prompted us to explore this idea to help communities prepare for the disaster before it strikes. Affected upstream communities living by river banks could convey messages that would be instantly shared with communities living downstream.
We weren’t wedded to the modality of how this information was going to be delivered, but the purpose and conviction that this needed to be done was not something we could abandon, despite the discouragement.
We had earlier reached out to a trusted and valued group of volunteers in the country, the Bangladesh Scouts, who are engaged in community development work. We wanted to meet with the Scouts who are themselves affected by the flash floods.
In the first week of August, the prototype design team spent a weekend with the Scouts. It turned out that many came from flash-flood zones, and they identified with our purpose and wanted to work with us. Their solution was to make fellow Scouts-members focal points for disseminating the information and they would use Facebook to alert each other.
In addition, they wanted training on how to interpret the data as to what it meant for the community when the rainfall predicts a two-metre rise in water levels - this interpretation has to be localized to be of use to the flash flood prone community.
In addition, we came up with a powerful idea, that has immense potential to be replicated across the flood-prone country. The loudspeakers that sit atop the tallest dome of every mosque could be used to announce warnings and real-time information.
This way people would have enough lead-time to probably relocate to safer areas, and perhaps protect their livelihoods, a big part of which is farming fish and growing rice. Nets could be used to save the fish from being swept away and, if enough time was available, standing crops could also be harvested if ready.
In trying to get a deeper understanding of what technology could do to help crowd source the information, we spoke to a number of other groups. The Petajarta group from Indonesia are using crowdsourcing for urban floods through twitter. In Sweden, crowdsourcing is being used to report on forest fires. All of these initiatives are not without challenges as crowdsourcing also comes with the inherent risk of how credible the information is and when the warning becomes critical enough to take positive action?
Now, our next step is to take this prototype to two communities, one upstream and one downstream along the riverbanks. The elements of the prototype include role of communities during normal times, crowdsourcing the warning when the flash flood is imminent, once it has hit the upstream communities and the as flood moves downstream. Once done, we believe our prototype will be ready for use.
Even three hours of lead time is very critical in case of flash floods and 36 hours could help save hundreds of lives every year. This would be the first time ever that someone has worked on the idea of upstream communities warning downstream communities using real time information and technology.
We’ve learnt something bigger through all of this. When you have conviction and purpose, pursue them. Failure and dismissal don’t stand up to persistence and belief.