Recently, the regions around the Dead Sea in Jordan were flooded, causing the death of 21 children who were on a school trip, and injuring 35 more. Such disasters affect millions of people every year and cause property damage worth hundreds of billions. In 2017 alone, almost 335 natural disasters have affected more than 95.6 million people, and killed 9,697, costing around US $335 billion.
But, the impact of these phenomena can be reduced if we were able to predict their occurrence. AI-powered systems can already predict the prices of stocks, which involve the analysis of numerous variables. Likewise, researchers are applying artificial intelligence to accurately predict natural disasters. By predicting the occurrence of natural disasters, we can save thousands of lives and take appropriate measures to reduce property damage.
Using AI to Predict Natural Disasters
Artificial intelligence has been helping us in various applications such as customer service, trading and healthcare. And now, researchers have found that AI can be used to predict natural disasters. With enormous amounts of good quality datasets, AI can predict the occurrence of numerous natural disasters, which can be the difference between life and death for thousands of people. Some of the natural disasters that can be predicted by AI are:
Researchers are collecting enormous amounts of seismic data for analysis using deep learning systems. Artificial intelligence can use seismic data to analyze the magnitude and patterns of earthquakes. Such data can prove beneficial to predict the occurrence of earthquakes. For example, Google and Harvard are developing an AI system that can predict the aftershocks of an earthquake. Scientists have studied more than 131,000 earthquakes and aftershocks to build a neural network. The researchers tested the neural network on 30,000 events, and the system predicted the aftershock locations more precisely when compared to traditional methods.
Similarly, multiple researchers are creating their own applications to predict earthquakes and aftershocks. In the future, we may be able to foresee earthquakes and authorities can start evacuation operations accordingly. Currently, Japan is using satellites to analyze images of the earth to predict natural disasters. AI-based systems look for changes in the images to predict the risk of disasters such as earthquakes and tsunamis. Moreover, these systems also monitor aging infrastructure. Artificial intelligence systems can detect deformations in structures, which can be used to reduce the damage caused by collapsing buildings and bridges, or subsiding roads.
Google is building an AI platform to predict floods in India and warn users via Google Maps and Google Search. The data for training the AI system is collected with the help of rainfall records and flood simulations. Similarly, researchers are developing AI-based systems that can learn from rainfall and climate records and tested with flood simulations, which can predict floods better than the traditional systems. Alternatively, AI can also be used to monitor urban flooding. Researchers at the University of Dundee in the United Kingdom are monitoring urban flooding by collecting crowd-sourced data with Twitter and other mobile apps. The data contains images and information about the location and situations in a locality, which is recognized by the AI. Such systems can be used to monitor and predict the damage done by floods along with other methods. Likewise, applications based on artificial intelligence and deep learning is useful for disaster management.
IBM is developing Watson that will predict volcanic eruptions using seismic sensors and geological data. IBM is aiming to forecast the locations and the intensity of eruptions with the help of Watson. Such applications can help to prevent the loss of life in areas surrounding active volcanoes.
Recently, NASA and Development Seed tracked Hurricane Harvey using satellite images and machine learning. The method proved to be six times better than the usual techniques, as the hurricane can be tracked every hour instead of every six hours with the traditional methods. Therefore, the developments in technology are helping in monitoring hurricanes and foreseeing the path of hurricanes, which can assist in mitigation efforts.
Realizing the Limitations of AI
Although artificial intelligence is advanced enough to help in various applications, the limitations of AI are preventing the use of technology in real-world applications. One of the limitations of artificial intelligence is that AI can compete with humans in the volume and speed of operations, but not in the quality of prediction. There are various instances where AI can make mistakes. The data fed to the system is collected by humans, which, can be flawed. Hence, the results produced by AI might be inaccurate. Additionally, looking at the previous trends in the adoption of AI, we have a tendency to be too dependent on AI. Thus, if AI makes a mistake, we would still trust the system and take actions accordingly. For example, if the system that predicts earthquake aftershocks makes an error in predicting the location of the aftershock, countless lives will be lost. Therefore, researchers still have to conduct multiple tests to make the technology reliable and fit for real-life implementation.
Another issue with artificial intelligence is that the data is based on past records of natural disasters. Therefore, AI-powered applications cannot process the changing trends and magnitudes of natural disasters such as floods and earthquakes. And, there is no way to introduce the effect of climate change on natural disasters in current AI-powered applications. Since AI is trained using past records, the systems are unable to analyze the effects of climate change. Hence, it is difficult for artificial intelligence to predict long-term trends of various natural disasters that are affected by climate change.
Developing a Road Map for Deploying AI
The introduction of modern technologies like artificial intelligence and machine learning is going to be useful to predict natural disasters. But, it is essential to tackle the limitations of technology before deploying AI in real-world applications. Hence, researchers need to focus on solving the existing issues with artificial intelligence.
To successfully deploy AI, government organizations need a roadmap that can simplify the process of adoption. The roadmap for effective adoption and application contains the following steps:
- Hire experienced researchers and tech experts who have worked with AI
- Collect good quality data for training the AI-powered application
- Enlist skilled professionals that can help in creating adoption strategies
- Update current in the government organization
- Educate government personnel about artificial intelligence
The adoption of AI to predict natural disasters will save millions of lives. Additionally, the datasets analyzed by the AI-powered systems will help in understanding the magnitude and the patterns of natural disasters such as floods, earthquakes and tsunamis, which can help in better planning of infrastructure in disaster-prone areas. Thus, government organizations need to deploy AI to predict natural disasters and monitor them accurately to ensure the safety of their citizens.
Naveen Joshi is Founder and CEO of Allerin, which develops engineering and technology solutions focused on optimal customer experiences. Naveen works in AI, Big Data, IoT and Blockchain. An influencer with a half a million followers, he is a highly seasoned professional with more than 20 years of comprehensive experience in customizing open source products for cost optimizations of large scale IT deployment.