Blog

Made in the Gulf: a new era of humanitarian aid

Patrick Meier is director of social innovation at the Qatar Computing Research Institute, which develops next generation humanitarian technologies. Credit: Kris Krug

Patrick Meier is director of social innovation at the Qatar Computing Research Institute, which develops next generation humanitarian technologies. Credit: Kris Krug

Big data will be a force for change, if global aid agencies can leverage technology to unleash its power. Patrick Meier reveals how Qatar is leading the charge.

Aid agencies today face a data overload. For an industry once marked by information scarcity, the problem is no longer finding the data, but sifting through to find what is relevant. Social media, mainstream media, satellite and increasingly aerial imagery, are just some of the drivers of data change.

For relief organisations, the deluge is akin to looking for a needle in a haystack. The ‘needle’ is potentially lifesaving information. The ‘haystack’ is the flood of other data generated during disasters. At the Qatar Computing Research Institute (QCRI), we believe the answer lies in next generation humanitarian technologies.

QCRI’s social innovation programme seeks to apply advanced computing for social good. A principal area of focus is data change in the humanitarian space. To this end, we work directly with leading aid organisations around the globe to help them win the big data battle.

In response to the Nepal earthquake in April, the United Nations used our MicroMappers platform to rapidly assess disaster damage. MicroMappers pairs crowdsourcing – in the form of digital volunteers from around the world – with artificial intelligence (AI) to sift through data created during crises. In the case of Cyclone Pam in March of this year, the UN used MicroMappers to quickly find and map more than 1,000 pictures of disaster damage from both social media and mainstream media.

AI plays a critical role in MicroMappers’ success. As digital volunteers find relevant tweets, our AI engine can be trained to learn from users in real-time. The technology reads all tweets tagged as ‘urgent needs’, and learns to recognise patterns. Once our AI engine has read enough, it will automatically tag new tweets – at a rate of up to 2 million per hour.

Still, many people in disaster-affected areas are not on Twitter, which is why we developed a system to detect and tag SMS messages. Our partners Unicef and the World Food Programme (WFP) are piloting this extension for public and emergency health projects in Africa.

Our team is now working to extend this real-time machine learning, to analyse pictures posted on social and mainstream media. We hope to develop automated algorithms that can identify disaster damage in Instagram pictures and YouTube videos, for example. This is no trivial task, and underscores why aid agencies are unable to combat data change alone.

Aerial data is fast becoming a big challenge too. At the time of writing, the World Bank is using MicroMappers to help tackle the aftermath of Cyclone Pam. The agency is using remote control helicopters and planes to capture aerial shots of disaster-affected areas, and utilising MicroMappers to filter them. The platform has previously been used by wildlife protection experts in Africa to analyse aerial imagery of their wildlife reserve for monitoring and protection purposes. Clearly, our humanitarian technologies can be used for much more than sudden onset disasters.

Looking to the future, the sky may not be the limit for next generation humanitarian technologies. We are currently collaborating with US-based Planet Labs to accelerate disaster damage analysis using satellite imagery. Planet Labs plans to deploy hundreds of micro satellites into space to create the first global constellation of satellites. This will give us access to daily imagery updates for the entire planet. We plan to use MicroMappers, among other platforms, to quickly crowdsource the analysis of satellite imagery for a wide range of humanitarian and social good projects. We also plan to use AI to create the algorithms that can automatically identify features of interest in satellite imagery.

The end goal is to have one map that pulls in all the filtered information from tweets, pictures, videos and imagery. Ultimately, we hope to transform the way aid agencies respond to global crises and to help solve major humanitarian challenges.

About the writer

Patrick Meier is an expert in innovation in aid and author of the book ‘Digital Humanitarians’. He is director of social innovation at the Qatar Computing Research Institute where he develops next generation humanitarian technologies.