When disaster strikes, AI helps responders reach people more quickly

Turning information into action

For the United Nations Office for the Coordination of Humanitarian Affairs (OCHA), quickly turning information into action can make the difference between reaching survivors and arriving too late.

"You're saving lives by providing this data to the team who is doing search and rescue, whose job is to bring out people who are under the rubble and might be alive," says Fawad Hussain Syed, a humanitarian affairs officer with OCHA in Geneva. "What this data is doing is basically getting to those teams in time so they can go with their dogs and machinery and find people who are living and bring them out."

After disasters, information often arrives in fragments through messages, news reports, social-media posts and scattered field observations that can be difficult to verify, Syed says. Satellite imagery with analysis provides critical evidence.

"It translates into a response plan," he says.

OCHA received HASTE's Venezuela damage analysis and rapidly distributed it through a humanitarian network used by thousands of people from governments, nonprofits, local authorities and emergency-response teams. That helped organizations begin building a shared understanding of where damage appeared most severe so they could estimate needs, plan operations and communicate the scale of the disaster to partners and donors.

Syed says that's "very, very crucial" for coordinating search-and-rescue responses, in particular, as teams arrive from all over the world to help - sometimes 10 to 15 of them, depending on the disaster.

"They land at the airport and the first thing they want to know is where to go," he says. "This is exactly what this data then provides."

Understanding the full scope of a disaster

Not long after the earthquake, responders faced a new challenge.

Groups were getting overwhelmed with imagery and analyses from multiple sources "flooding the system," says Zachary Arno, a data scientist with OCHA. He quickly built a web viewer, hosted on Azure, that brought HASTE assessments and eight other disaster-mapping products together in one place so organizations could compare them and build a more complete picture of what they were all showing about conditions on the ground.

The Microsoft data arrived with detailed metadata and confidence rankings, Arno says, making it easier to incorporate into that broader analytical picture.

"Speed and communication and transparency helped shape the viewer," he says.