Scientists Have Developed a Device for the Search for Hydrocarbons Based on Artificial Intelligence
Specialists from Tyumen State University, together with the Institute of Geophysics of the Ural Branch of the Russian Academy of Sciences and a manufacturer of geophysical equipment and oilfield equipment, have developed a neuromorphic (that is, simulating the activity of the human brain) device. It performs an acoustic search for oil and gas rich objects thanks to the built-in algorithm for recognizing the sounds of an oil well, which allows you to increase production.
As told in the Tyumen State University, the recognition process occurs due to the neuroprocessor, which scientists create in the laboratory of micro- and nanoelectronics of the Center for Nature-Inspired Engineering. The mechanism works by analogy with the neural networks of living organisms, that is, it encodes information based on input data. Such neuroprocessors are used for computer vision systems, machine learning.
If earlier innovations were created from scratch, now scientists are trying to recreate the unique properties of natural objects, explains Alexander Pisarev, Leading Researcher at the Laboratory of Nanomaterials and Nanoelectronics. The experience of Tyumen people in the implementation of artificial intelligence tools for the digitalization of the industry, and not only the fuel and energy complex, has already become interesting in specialized companies.
The developer’s device will carry out wave action and neural network analysis of seismoacoustic noise in the rock. With the help of this technology, fuel and energy companies can more effectively cope with production tasks. For example, to increase the oil recovery factor, increase the resource recovery in the productive interval of the reservoir and determine the sources of watering.
Also, the mechanism can be used to perform geophysical work in wells of complex design.
