How Machine Learning Can Help Predict Floods and Fires

Google and Amazon Web Services (AWS) have highlighted their respective work on machine learning (ML) models that can help countries cope with environmental crises that occur with increasing regularity around the world.

Companies have highlighted their efforts to tackle the effects of climate change, such as floods and forest fires, as COP26 in Glasgow has just ended.

The flood forecasting system developed by Google

Google published a separate article on its flood forecasting system with machine learning models that it says provides “accurate real-time alerts to agencies and the public, with a focus on flooding in large rivers.” The article was written by Google researchers. Research and the Hebrew University of Jerusalem in Israel.

Google’s flood forecasting initiative, launched in 2018, sends alerts to the smartphones of people living in flood-affected areas. She is part of Google’s Crisis Response Program, which works with frontline and emergency workers to develop technology.

Since 2018, the program has spread to much of India and Bangladesh, encompassing an area populated by some 220 million people. Since the 2021 monsoon, it has expanded further to cover an area where 360 ​​million people live. “Thanks to better flood forecasting technology, we’ve sent more than 115 million alerts, that’s about three times what we used to send,” says Yossi Matias, Google’s vice president of engineering and head of crisis response, in a blog post.

Google alerts don’t just show how many inches a river has flooded. Thanks to your new machine learning models that use long-term memory deep neural networks (LTSM), you can now provide “flood maps” that show the extent and depth of floods as a layer on Google Maps.

The researchers state that “the LSTM models performed better than the conceptual models that were calibrated on large data sets in each basin.”

“Although previous studies have provided encouraging results, it is rare to find real operating systems whose main components are ML models capable of calculating accurate and timely flood alerts,” said the Google researchers.

AWS works with AusNet on fires in Australia

Meanwhile, AWS worked with AusNet, an energy company based in Melbourne, Australia, to help mitigate wildfires in the region. AusNet has 54,000 kilometers of power lines that distribute power to approximately 1.5 million homes and businesses in the state of Victoria. It is estimated that 62% of the network is in high risk areas for forest fires.

AusNet used cars equipped with Google Maps-like LiDAR cameras and the Amazon SageMaker machine learning system to map areas of vegetation in the state that must be cut down to contain the risk of wildfires. Their previous system was based on a geographic information system and used custom tools to label LiDAR points.

AusNet worked with AWS to automate LiDAR point classification using AWS managed machine learning models, GPU instances, and S3 storage.

AusNet and AWS created a semantic segmentation model that accurately categorized 3D point cloud data for conductors, buildings, poles, vegetation, and other categories, AWS notes in a blog post. “The team was able to train a model at a speed of 10.8 minutes per epoch on 17.2 GB of uncompressed data across 1,571 files for a total of approximately 616 million points. For inference, the team was able to process 33.6 GB of uncompressed data in 15 files for a total of 1.2 billion points in 22.1 hours. This translates to an average inference of 15,760 points per second, including amortized startup time, ”said AWS.

“Being able to quickly and accurately label our aerial survey data is critical to minimizing the risk of wildfires,” says Daniel Pendlebury, Product Manager for AusNet. “By working with the Amazon Machine Learning Solutions lab, we were able to create a model that achieved an average accuracy of 80.53% on the labeling data. We look forward to reducing our manual labeling efforts by up to 80% with this new solution. “

Source: .com

Woodmart Theme Nulled, WP Reset Pro, Newspaper 11.2, Newspaper – News & WooCommerce WordPress Theme, Premium Addons for Elementor, Rank Math Seo Pro Weadown, WeaPlay, WordPress Theme, Plugins, PHP Script, Jannah Nulled, Elementor Pro Weadown, Woocommerce Custom Product Ad, Business Consulting Nulled, Jnews 8.1.0 Nulled, Avada 7.4 Nulled, Nulledfire, Dokan Pro Nulled, Yoast Nulled, Flatsome Nulled, PW WooCommerce Gift Cards Pro Nulled, Astra Pro Nulled, Woodmart Theme Nulled, Slider Revolution Nulled, Wordfence Premium Nulled, Elementor Pro Weadown, Wpml Nulled, Consulting 6.1.4 Nulled, Fs Poster Plugin Nulled

Back to top button