Russian Scientists Build AI System to Speed Up Mineral Exploration
Researchers at Moscow’s top physics institute have created an AI-powered tool that transforms seismic waves into underground maps up to 20 times faster, opening new frontiers for mining and construction safety.
A team at the Moscow Institute of Physics and Technology has unveiled a breakthrough method to accelerate the search for mineral deposits, improve Earth mapping, and assess seismic risks for construction.
The system relies on artificial intelligence to drastically speed up the conversion of seismic waves into detailed maps of underground structures. Traditional diffusion models require vast amounts of computation. By contrast, the new approach — called conditional Image-to-Image Schrödinger Bridge (cI2SB) — cuts the process down to just dozens of steps.
At its core lies the mathematics of diffusion, while the neural network works like a filter, gradually clarifying blurry data. Step by step, the AI adds detail, producing sharper geological maps.
The tool has already been tested on large synthetic datasets, proving almost 20 times more accurate and faster than earlier methods. Crucially, it can detect subtle geological features that older techniques often miss.
