The company's Disaster Mapping System uses satellite and drone images to help pinpoint buildings most impacted by a disaster such as a tornado or earthquake. The modeling is designed to help first responders.
(TNS) — Search and rescue operations after a natural disaster can be dangerous, difficult and dependent on highly trained volunteers and workers.
Artificial intelligence company Hypergiant Industries says it has developed a product that could make the process better -- an AI model that uses aerial images to prioritize disaster rescue operations.
The company's Disaster Mapping System uses satellite and drone images to help pinpoint buildings most impacted by a disaster such as a tornado or earthquake. The modeling is designed to help first responders understand which areas have been hardest hit.
Texas-based Hypergiant Industries, which has about 195 employees in Austin, Dallas, Houston, Washington D.C. and Seattle, focuses on how to solve problems using artificial intelligence and emerging technology. Its technology is used in space science and exploration, satellite communications, aviation, defense, health care, transportation and municipal infrastructure, as well as the food and beverage industry.
"With the growing threat of death and destruction, we just wanted to support efforts" Hypergiant founder Ben Lamm, said.
The AI model gives rescuers a birds-eye view of areas affected, and detects changes in an environment such as a fallen structure or a caved-in roof.
"It can be hard to know where to start," Lamm said. "You can give (a first responder) something like a map that's color-coded to say here are the areas of where they need to look first, because they were absolutely just obliterated."
In 2018, 315 natural disaster events affected over 69 million people and led to 11,804 deaths, according to the Centre for Research on the Epidemiology of Disasters. These events caused $131.7 billion in economic losses across the world.
Anyone using the model can import their own satellite images or drone data to detect the extent to which an area has been damaged, and how severe that damage might be. The model uses machine learning and artificial intelligence to analyze changes in the images. It then takes the images and color codes the areas that received the most damage.
Hypergiant is open-sourcing the model and adding it and several projects to a marketplace technology consulting giant Booz Allen Hamilton is developing. The companies announced the formation of a partnership in July.
The marketplace, called Modzy, is an AI model focused store. It's designed to bridge companies and the government, with immediate access to scalable, trusted AI-driven models.
Lamm said the model is being open-sourced in hopes that others will be able to build on the model and make it more robust and increase capabilities to save lives.
©2019 Austin American-Statesman, Texas. Distributed by Tribune Content Agency, LLC.