AI will help calculate "nuclear terrorists": how does it work out
The rest of all attempts to create units in other countries or individual groups are terminated rigidly. The AI -based complex is intended to facilitate the task of monitoring the non -proliferation of nuclear threat. The task of the complex will include such features based on the chemical analysis of radioactive substances and the general analysis of the suitable infrastructure for the possible production of nuclear weapons.
According to PNNL Director Stephen Ashby, the neural network will help not only in identifying nuclear threats, but with the help of special methods will make it much faster than experts. AI will use a special car coding model in their work. It will process the image of radioactive material to determine its origin and method of manufacture. Special software marks samples and compares them to a database from universities and other national laboratories.
According to experts, each radioactive material has its own unique structure, a "imprint", which reports the place of production and the method of processing. Therefore, Russia would not be able to carry out nuclear diversion against Ukraine in order to accuse our country of use of uggage or radioactive waste, since All nuclear materials used by the Russian Federation are well known.
Depending on how similar these particles are already in the library, analysts can evaluate the purity of the unknown sample and trace the source materials to the possible laboratories of nuclear production. This is useful and important if you need to find out if good material for creating nuclear weapons and how far the creators have advanced in its manufacture, and how and where they get such material.