TopAgent IC – ROCKINSIGHT | Automating Core Sample and Geological Analysis with Sovereign AI

Learn more about the ROCKINSIGHT solution from TopAgent IC (in collaboration with AIQ) – applying artificial intelligence and computer vision technology on a sovereign AI platform in the energy sector to fully automate core analysis.

1. The Importance of Core Sample Analysis in Energy Exploration and Exploitation

In the value chain of the energy industry, accurately evaluating the reserves and production potential of an underground reservoir is a decisive step. To achieve this, geologists and engineers cannot rely solely on seismic data or indirect parameters. They need direct physical evidence, which comes in the form of core samples drilled and extracted from thousands of meters below the surface.

Particularly for limestone reservoir structures, the geological complexity is immense. Unlike sandstone, which has a relatively homogeneous structure, carbonate rocks often undergo intense geological alteration, creating an intricate, irregular network of pores and fractures. Therefore, studying thin-section petrography under electron microscopes to determine porosity, permeability, and mineralogical characteristics is an essential task for planning effective energy exploitation.

2. Bottlenecks of Traditional Geological Analysis Methods

Before digital technology intervened, the core analysis process was an arduous, expensive, and error-prone procedure. This required teams of experts to cut rock samples into thin slices and manually inspect them under polarizing microscopes.

This process had several key limitations:

• Over-reliance on human factor and subjective experience: Analysis results were often heavily influenced by the skill and perception of individual experts. Two different engineers could reach different conclusions from the same rock sample.
• Time-consuming and costly: Manually quantifying components took weeks. This created a bottleneck that delayed decision-making during expensive exploration drilling campaigns.
• Risk of domain expertise loss: When veteran geologists with decades of experience retired, their vast knowledge went with them. Paper reports or scattered files could not help the next generation inherit their image interpretation expertise.

3. ROCKINSIGHT Solution – Geological Analysis Powered by AI

To address these challenges, the ROCKINSIGHT solution leverages Computer Vision combined with Deep Learning to automate the end-to-end micro-image analysis of rock thin sections, specifically optimized for complex carbonate reservoirs.

Built upon the TopAgent IC Sovereign AI platform, ROCKINSIGHT ensures that micro-geological data—a strategic national asset—is processed and analyzed within a closed on-premise environment. This completely prevents data theft risks associated with sensitive industrial assets.

The ROCKINSIGHT system goes beyond basic image recognition, automatically providing 6 in-depth analytical capabilities (corresponding to 6 core features) for a single thin-section image:

• Original Image: The system digitizes and stores ultra-high-resolution thin-section images, allowing engineers to zoom, pan, and examine details digitally from anywhere.
• Predicted MICP (Mercury Injection Capillary Pressure): This AI in energy feature models and predicts MICP curves from 2D images, enabling fast determination of pore-throat size distribution without needing hazardous and expensive physical mercury injection lab tests.
• Depositional Environment: By analyzing microstructures, the AI automatically classifies and identifies the rock’s original depositional environment (e.g., shallow shoals, reefs, lagoons), helping reconstruct the geological history of the reservoir.
• Pore: The system automatically segments pore space from the solid rock matrix. It precisely calculates primary vs. secondary porosity percentages, highlighting fluid flow pathways.
• Cement: Identifies mineral precipitation areas that fill pore space (cementation). This step is critical because cementation reduces natural porosity and hinders production capabilities.
• Framework Grain: Automatically classifies and quantifies rock-forming grains (such as ooids, peloids, bioclasts). This serves as the foundation for rock classification according to international standards like Dunham or Folk.

4. Operational Value and Practical Implementation Efficiency

Deploying AI-driven geological analysis via ROCKINSIGHT delivers numerous advantages, transforming workflows across core labs in energy enterprises:

• Reduced Operational Costs (OPEX): By automating manual tasks, companies significantly cut outsourcing expenses and lower expensive dependencies on international physical labs.
• Faster Decision-Making: Automated image analysis yields instant, 100% consistent results. Engineers can immediately adjust drilling trajectories based on real-time geological updates rather than waiting months for lab reports.
• Knowledge Retention: ROCKINSIGHT acts as a permanent knowledge repository. AI algorithms preserve expert interpretation experience while eliminating human bias and subjectivity.
• Data Synchronization: All petrographic analysis results from microscopic images are digitized and unified with macroscopic geological data (such as seismic data and well logs). This integration eliminates communication silos across departments, optimizing team collaboration for reservoir management strategies.

As easily accessible energy reserves become depleted, decoding complex rock formations like carbonates is key to boosting production. Adopting the ROCKINSIGHT solution under the TopAgent IC Sovereign AI Platform to automate core analysis not only cuts costs but also reinforces a company’s technological leadership in the global energy digital transformation. To stay updated on other pioneering technological trends, explore our AI & Physical AI category.

Contact ADT for in-depth consultation and comprehensive documentation on Sovereign AI solutions in the Energy Sector!