In 2014, Stanford student structural engineer Ahmad Wani was visiting family in his native Kashmir when a catastrophic flood struck. The rising waters stranded him and his family for seven days without food or water, during which they watched their neighbor’s home collapse, killing everyone inside.

After this horrifying experience, Wani was struck by just how disorganized the emergency response was. “There is no science behind how people should be rescued,” he says. “Disaster response is really random.”

Today, Wani’s startup One Concern is launching a machine learning platform that provides cities with specialized maps to help emergency crews decide where to focus their efforts in a flood.

The maps update in real-time based on data about where water is flowing to estimate where people need help the most. It’s the latest in a wave of AI-powered tools aimed at helping cities prepare for an era of severe, and increasingly frequent, disasters.

Artificial intelligence, such as the platform One Concern has developed, offers a tantalizing solution. But it’s new and largely untested. And the urgency is growing by the day.

Wani, Hu, and Frank started One Concern in 2015 and then released its earthquake platform, called Seismic Concern, in 2016. Seismic Concern predicts the damage caused by earthquakes on a block-by-block level and is now used by eight different municipalities, including the cities of San Francisco, Los Angeles, and Cupertino.

Now, the company is launching Flood Concern, a constantly evolving risk map that crunches huge amounts of data based on the physics of how water flows, information about previous floods, and even satellite imagery to approximate the depth, direction, and speed of the water–and determine which areas of a city are most at risk.

Read the full article about AI for disaster relief by Katharine Schwab at Fast Company.