Waze for Disasters

Sup? 

The other day, I caught up with my friend Erich Broksas. He's the Head of Strategy at World Central Kitchen (WCK).

WCK is the brainchild of José Andrés (yes, that José Andrés). It’s a non-profit that partners with chefs to feed populations affected by disasters.

WCK has responded with support after dozens of natural disasters, including Hurricane Maria in Puerto Rico, the 2010 Haitian earthquake, and, most recently, the humanitarian crisis caused by the Venezuelan government. For his work, José was recently nominated for the Nobel Peace Prize. 

Erich responded to Hurricane Michael in Panama City, Florida. WCK set up a makeshift kitchen run by a local restaurateur.

A constant stream of volunteers and resources flowed in and out of the area during the response effort. But there was an issue: the hurricane made it more difficult to find the kitchen’s exact location (especially for responders who were non-residents unfamiliar with the area). So how did they navigate? Not with their phones. Instead, they did what humans have been doing for millennia—used landmarks 

From Erich: "It was, go to the stalled Mercedes at the intersection and turn left...."

So, if that Mercedes moved, it got even harder to get where they were going.

Navigation apps like Google Maps, Waze, and Apple Maps are great for our day-to-day lives but lack some core functionality critical to disaster zones.

Waze Failz

When Waze navigates me to a closed road along my route home, it's annoying, but not the end of the world (it’s happened three times this year, not that I’m counting). If it occurred during a natural disaster, though, I could miss something critical, like downed wires or a flooded road.

During the California wildfires, Waze told drivers to head down I-405 in LA, despite the fact that those were the neighborhoods literally on fire.

During Hurricane Florence last year, many users following Waze's directions were navigated to flooded roads. Oops.

In this environment, the human-eye often beats the machine. But, perhaps there is a way for the machine to catch back up?

Waze + Computer Vision + AR for disasters

Enter today's idea: a version of Waze or Google Maps designed for disasters.

As a user navigates around a disaster zone, the app would use GPS to track his or her location. Simple enough. We'd then layer in computer vision and augmented reality to contextualize the environment.

When a user fires up the app, it'd look a bit like Google Maps' new interface. Not only can you navigate, but you can also see landmarks right in front of you, in context.

As the user moves his or her phone, it would gather information about the surroundings. Users could get navigation information specific to their exact location. And, it would be more accurate than GPS, which is often off by up to several meters.

This scan could gather data about users' surroundings and update others about a rapidly-changing environment. For instance, if the scan records toppled utility pole, it would mark the area unsafe for others.

The user of the app could also tag critical information (e.g. collapsed buildings or unsafe bridges.). In a rapidly changing environment, this could prove critical to the safety of others

Relief organizations like WCK, local emergency management, and FEMA could use this to monitor their environments in real-time.

Imagine a war room, where everyone sees a topographically-accurate map and drills into specific micro-regions for context.

What do you think? Would you download a disaster app?

Special thanks to my new insider group, especially Dan Krawec for their input on this piece. If you are interested in contributing to these posts before they're shared, reply to this email.