In a recent study published in Nature Scientific Reports, the University Of Tokyo Institute Of Industrial Science, has introduced a new AI tool that can predict the spread of nuclear fallout from nuclear disasters.

The AI, as detailed in the report, was trained on the data set of weather-related data that has been gathered over the years. This is important to then demonstrate to the machine all the contributing variables in the program. Even when making the predictions in real-world situations, the program will use real-time reports on weather forecasts and wind patterns.

Evidently, there are many factors involved in the complex algorithm of this program that the team promises to share upon ‘reasonable request’. The lead author of the report, Takao Yoshikane, had this to say about the matter “Our new tool was first trained using years of weather-related data to predict where radioactivity would be distributed if it were released from a particular point. In subsequent testing, it could predict the direction of dispersion with at least 85% accuracy, with this rising to 95% in winter when there are more predictable weather patterns.”

With such high levels of accuracy, the program can actually be relied upon when framing the evacuation procedures for nuclear accidents. The report suggests that it can offer 30 to 33 hours of grace period to the authorities, to implement proper evacuation arrangements in the most susceptible areas.

Here’s a demonstration video that shows how the AI predicts the Distribution of Radioactive Material Deposition-

How the Technology Affects Us?

The present state of Chernobyl is a current testimony of how disastrous nuclear catastrophes can be. Not only can they impact the natural rhythm of human lives, but can also render a place uninhabitable for decades. Therefore, any technology or invention that can prevent such disasters or, at least, reduce the consequences of such accidents, brings true relief to society.

In any case, such a technology was in demand for quite some time. Dispersion of air pollutants has always remained an inaccurate science for the geoscientists which, in turn, has always limited the scope of their solutions in rescue efforts.

Now, when there is a system that can accurately predict the dispersion of nuclear fallout, that can do so with minimal manual effort from humans, the rescuers will have better premises to strategise the evacuation processes and even contain the spread of harmful releases.

What’s Next?

AI is progressing at a rapid pace to bring us to the levels of the great complexity frontier, but the real utility of the technology only comes from these newer advanced technologies. We’ve witnessed how the lack of suitable atmospheric modelling tools has created global hysteria during the second most severe nuclear power plant accident, the Fukushima Daiichi nuclear disaster on 11 March 2011. There were no reliable analytical systems available to guide the counterbalance measures, leading the World Meteorological Organization and the United Nations Scientific Committee, on the Effects of Atomic Radiation to produce five different atmospheric transport models of the incident’s radionuclide dispersion. These models, since then, were in waiting for major improvements for better application in future cases, and this study indeed proves to be one of these models.

Resources:

https://www.nature.com/articles/s41598-018-27955-4

https://www.iis.u-tokyo.ac.jp/en/news/2932/

https://interestingengineering.com/new-ai-can-predict-the-distribution-of-radioactive-fallout-during-nuclear-disasters

https://www.artificialintelligence-news.com/2018/06/25/embracing-the-power-of-ai-the-process-behind-starting-an-ai-project/