Early warnings for flood detection require satellite
communication, but there’s a bottleneck in receiving.
Data transmission to terrestrial receivers is limited.
World Floods, an AI-based model for flood detection,
accelerates response times by processing relevant
information on a device, reducing the amount of data
WorldFloods, an AI-based model for flood detection, accelerates
response times by processing relevant information on a device,
reducing the amount of data needing transmission.
Early warnings for flood detection require satellite communication,
but there’s a bottleneck in receiving. Data transmission to
terrestrial receivers is limited, which is something researchers
hope to remedy in a new artificial intelligence project. A new nano-satellite
outfitted with an artificial intelligence model will provide flood
monitoring that hopefully reduces response times through neural
networks and on-device processing.
Remote monitoring for flood detection
Researchers at the Image Processing Laboratory (IPL) of the University
of Valencia, in collaboration with the University of Oxford and the
Phi-Lab of the European Space Agency (ESA), developed WorldFloods, a
model for flood detection using a neural network. The model will
accelerate response times by processing relevant information on the
device, reducing the amount of data needing transmission.
This process ushers in near real-time terrain scanning and reduces
the computational load required of current detection systems. It creates
more compatible transfers between the large images’ sensors produce and
the smaller devices required for monitoring. The rapid exchange of
information could usher in a new era of environmental monitoring.
New nano satellite launched
CubeSats are part of the new arrangement. They help reduce something
called revisitation time, or the time it takes for a satellite to
recover an area previously noted. Along with five other CubeSats,
the WorldFloods CubeSat launched in June from Cape Canaveral and
carried a first-of-its-kind in-orbit processing service with
radiation-resistant chips. Researchers implanted the chips with
the new flood detection model and will use this to monitor conditions
more closely and gain more lead time to respond to changes.
The WorldFlood model was developed at the ESD-funded FDL-Europe
research center. FDL applies artificial intelligence to the pursue
of challenges in space, including disaster response, climate change
monitoring, and space weather or threats. It offers computing
resources designed for rapid iterations.
Provided by the IKCEST Disaster Risk Reduction Knowledge Service System