The U.S. Department of Energy, along with Microsoft, announced a partnership to develop artificial intelligence tools to help rescuers better respond to fast-changing natural events, such as floods, fires and hurricanes.
The First Five consortium (the name emphasizes the importance of the first 5 minutes in responding to a natural disaster) wants to build several dozen different systems based on AI. The development and testing works will be managed by the Department of Energy and Technology. In turn, Microsoft will provide technological resources, including the Azure cloud, for training and analysis in the field of the artificial intelligence model. Also other organizations, including public and private sector entities, are expected to participate.
Here are just so many technologies where we can solve some of the toughest problems, in a moment where we’re having an explosion of wildfires and floods and some really major natural disasters – says Cheryl Ingstad, director of the Energy Department’s Artificial Intelligence and Technology Office. Indeed, e.x. California faces another summer of wildfires, while Iowa grapples with devastating winds.
The rescue may come in the fall, when the consortium plans to release the first prototype. In turn, in the long term, it is planned to create a number of artificial intelligence systems, that can help save lives, property and resources. We can bring AI to bear here and help save lives – says Cheryl Ingstad.
The consortium already has two systems, that are at an early stage: one for mapping and predicting so-called fire lines or the bounds of an active wildfire, and a similar tool for floods. In each one, artificial intelligence plays a role, analysing huge and diverse datasets – such as fire temperatures or wind direction – in real time, which can help rescuers better allocate resources or inform the public.
The group will use the Microsoft Azure cloud to store various data elements, such as animations, showing the ‘behaviour’ of previous fires. The platform also includes pre-trained AI models, e.g. for object detection.
First responders, including those fighting fires, often lack real-time information, says Chris Renschler, an associate professor of geography who studies disaster management at the University at Buffalo – Having systems that can provide real-time information and even simulate real-world scenarios can make a difference. With simulation, if you can make more realistic scenarios, you can plan better for the extremes. It’s not only being able to better respond, but also having realistic scenarios out there that we can prepare and train for.