Close Menu

    Subscribe to Updates

    Get the latest creative news from FooBar about art, design and business.

    What's Hot

    FAO urges stricter checks on recycled food packaging

    May 15, 2026

    Hantavirus contacts in France and Netherlands test negative

    May 15, 2026

    Russian AI patent streamlines geological core analysis

    May 15, 2026
    Lloyds PostLloyds Post
    • Home
    • Contact Us
    • Automotive
    • Business
    • Entertainment
    • Health
    • Lifestyle
    • Luxury
    • News
    • Sports
    • Technology
    • Travel
    Lloyds PostLloyds Post
    Home » Russian AI patent streamlines geological core analysis
    Technology

    Russian AI patent streamlines geological core analysis

    May 15, 2026
    Facebook WhatsApp Twitter Pinterest LinkedIn Telegram Tumblr Email Reddit VKontakte

    INNOPOLIS, RUSSIA / EuroWire / — Russian researchers at Innopolis University have secured a patent for an artificial intelligence method that analyzes photographs of drilling core to identify fractures, faults, veins, breccias and other geological structures, a step designed to speed up rock classification and geological modeling. The patent, RU2856857C1, was published on February 25, 2026, and names Innopolis University as the assignee. It covers a method for clustering core image data for structural-lithological classification of rocks recovered during exploratory drilling.

    Russian AI patent streamlines geological core analysis
    Patented core imaging software highlights expanding AI use in geological data analysis. (AI-generated image for representative purposes)

    The invention lists Ilmir Nugmanov, Arseny Pinigin, Artur Shagitov and Aikhem Bouabid as inventors. According to university materials released in May, the project addresses one of the most labor-intensive parts of subsurface analysis, the manual description of core boxes used to document the composition and structure of rock from drilled intervals. Core photographs are widely used in exploration and mine planning because they preserve a visual record of fractures, bedding, mineral veins and other features that can influence interpretation of a deposit.

    The patented workflow uses a two-stage image processing system. In the first stage, a transformer-based neural network scans photographs of core boxes, isolates meter-long core sections and links each section to the correct depth interval. In the second stage, a pre-trained semantic segmentation model analyzes each section to detect structures including cracks, destroyed zones, faults, veins, breccias and streaks. The patent abstract also says the method removes technogenic cracks from the target area before post-processing and calculation of additional geological features.

    Patent details and workflow

    For each section, the system generates what the university describes as a digital fingerprint containing 2,780 numerical values per image. Those values include indicators tied to texture, color, contrast and the presence of fractures. The algorithm then clusters multidimensional feature vectors to group similar structures and highlight anomalies in the rock record. By converting photographs into structured data linked to depth coordinates, the method is designed to support more consistent classification of core images across large volumes of material.

    Innopolis University said the system classified core photographs in the same way as an experienced geologist in about seven out of 10 cases during its reported testing. University materials describe the work as part of efforts to reduce the time and subjectivity involved in manual core documentation. Arseny Pinigin and Ilmir Nugmanov, both identified by the university with roles in its oil and gas technology work, are among the named inventors on the patented method.

    Applications in exploration and construction

    The development is aimed at tasks in geological exploration, mining and construction, where rapid assessment of rock structure can affect decisions on deposits, wells, quarries and engineering conditions. The university said the clustering approach is particularly useful for identifying complex faults, tectonic breccias and other anomalous structures that can influence stability assessments. Because the system maps each detected feature to a depth interval, it can organize image-based observations in a format that can be used alongside broader geological interpretation.

    The patent was filed on May 27, 2025, and its publication in February 2026 formalized the legal protection for the method in Russia. Together with the university’s May 2026 technical description, the grant outlines a workflow that combines automated section detection, semantic segmentation and clustering for rock core analysis. For Innopolis University, the result is a patented AI tool focused on turning core box photographs into classified geological data without relying solely on manual inspection.

    Related Posts

    FAO urges stricter checks on recycled food packaging

    May 15, 2026

    Hantavirus contacts in France and Netherlands test negative

    May 15, 2026

    Putin says Russia tests Sarmat missile successfully

    May 14, 2026

    WHO says Andes virus spread occurred aboard MV Hondius

    May 14, 2026

    Manchester City Women open £10m first-team base

    May 14, 2026

    EU seals vital medicines rules to curb shortages

    May 13, 2026

    Editor's Pick

    FAO urges stricter checks on recycled food packaging

    May 15, 2026

    Hantavirus contacts in France and Netherlands test negative

    May 15, 2026

    Russian AI patent streamlines geological core analysis

    May 15, 2026

    Putin says Russia tests Sarmat missile successfully

    May 14, 2026

    WHO says Andes virus spread occurred aboard MV Hondius

    May 14, 2026

    Manchester City Women open £10m first-team base

    May 14, 2026

    EU seals vital medicines rules to curb shortages

    May 13, 2026

    EU steps up cruise ship hantavirus outbreak response

    May 13, 2026
    © 2024 Lloyds Post | All Rights Reserved
    • Home
    • Contact Us

    Type above and press Enter to search. Press Esc to cancel.