Helping Communities Better Predict Natural Disasters
Google’s Groundsource
Google introduced Groundsource, a Gemini-powered methodology that transforms public information into a high-quality record of historical disaster data, starting with flash floods in urban areas. Using Gemini, Groundsource analyzed decades of public reports and identified over 2.6 million historical flood events spanning more than 150 countries. With this dataset, researchers trained a new model that makes tangible progress toward predicting flash floods in urban areas up to 24 hours in advance. Google is using this technology to build a more resilient future and continue its work to ensure no one is surprised by a natural disaster.
Identifying Network Damage After a Storm
Verizon Digital Twin technology
Verizon’s Digital Twin technology is a revolutionary virtual modeling system that is changing the face of emergency restoration and accelerating storm recovery efforts. It uses drone-captured 3D imagery and AI to instantly identify the exact location of network damage after a storm, allowing engineers to prioritize repairs and enhance safety by reducing the need for on-site visual inspections. By comparing “before” and “after” imagery of the same cell sites and equipment, advanced software can instantly identify the exact location of any damage to antenna, cables, or other equipment with pinpoint accuracy, and allow engineers to identify and prepare for repairs without tower climbs and field visits.
Collecting and Distributing Donations During Disaster Response Efforts
Clara, The American Red Cross Chatbot
The American Red Cross’s AI agent, Clara, built with an Amazon Web Services grant, equips those who need help and those who would like to give help with access to essential information instantly. Clara can answer questions about disaster response, like providing up-to-date information on shelters open in your area, scheduling blood donations, locating where to give and receive donations, and more. According to one Amazon leader, agents like Clara can create a common operating picture of what is happening on the ground so that disaster relief teams can ask good questions and get help where it is needed most.