We are thrilled to launch our new Bayesian Facies Classification module for OpendTect! This tool enables the estimation of the most likely facies model along with probability maps for each facies.

Given a pair of seismic attributes and a set of well logs, facies are estimated based on the Bayesian posterior probability. This probability is defined by the product of a prior facies proportion and the bivariate conditional probability of the attributes given the facies, which is derived from the well logs. The most likely facies model is obtained by identifying the facies with the maximum posterior probability.

Key Features:
We designed an interactive interface to make this advanced workflow efficient and user-friendly:
📊 From Logs to Probability: Automatically compute Probability Density Functions (PDFs) for seismic attributes directly from your well logs.
🖱️ Interactive Crossplots: Visually manipulate PDFs to refine your facies definitions on the fly.
⚡ Instant Results: Experience instantaneous classification updates as you adjust parameters, perfect for testing different scenarios.
✅ Easy QC: Seamlessly validate predicted facies probabilities against reference data to ensure confidence in your interpretation.

Level up your quantitative interpretation! 👇

Interest in a trial for the LIS plugin? Send a message to Leandro Passos de Figueiredo and get started.
Don’t forget to take advantage of our release discount available for a limited time! 💸

#Geophysics #SeismicInversion #Bayesian #ReservoirCharacterization #QuantitativeInterpretation

We are thrilled to announce the official release of the Multi-Trace Wavelet Estimation module, now fully integrated into the LTrace Inversion Suite (LIS) plugin for OpendTect!

This module offers an intuitive, interactive interface designed to help you accurately define wavelet phase and energy by leveraging statistical analysis across multiple seismic traces and wells simultaneously.

🔍 What’s new in this module?
- Multi-Well Integration: Perform statistical analysis of multiple traces around several wells at once for a more robust estimation.
- Smart Correlation: Automatically compute the optimal time shift for each trace to achieve maximum correlation with your synthetics.
- Interactive Selection: A user-friendly interface that lets you manually pick and refine the traces used in your workflow.
- Fine-Tuned Accuracy: * Phase Tuning: Determine the absolute best wavelet phase based on your selected data.
- Energy Tuning: Once the phase is set, pinpoint the ideal energy value for high-fidelity results.

💡 Ready to elevate your inversion workflow?

Interest in a trial for the LIS plugin? Send a message to Leandro Passos de Figueiredo and get started.
Don’t forget to take advantage of our release discount available for a limited time! 💸

More info: https://ltrace.com.br/lis

Special thanks to Daniel Maia from BWE for building it up together

#SeismicInterpretation #SeismicInversion #Geophysics #WaveletEstimation #Exploration #Geosciences

As we wrap up 2025, it’s time to look back at the incredible journey of GeoSlicer, our open-source platform dedicated to digital rock physics (https://lnkd.in/dgjfxrgb). This year has been defined by massive technical leaps, a commitment to open science, and an ever-growing community of contributors.

From transitioning to a more stable build to expanding our Pore Network Modeling capabilities, here is the breakdown of what we achieved this year:
➡️ Version Leap: Moved from v2.5.3 at the end of 2024 to our latest stable release, v2.8.2.
➡️ Codebase Evolution: 1,070 commits with over 425,000 lines of code added.
➡️ Community Driven: 16 unique contributors pushing the boundaries of what’s possible in rock physics.
➡️ Stability First: Estimated 300 bug fixes and 172 brand-new features implemented.
➡️ Key Highlights & Innovations
➡️ Pore Network Modeling (PNM): Enhanced simulation for k_abs and k_rel calculations, improved handling of wettability, and new support for remote execution.
➡️ Advanced Segmentation: Introduced the "Smart Foreground" effect and Euclidean Distance Transform to make image analysis more precise than ever.
➡️ MicroCT & ImageLog: New data loaders (NetCDF/Tescan file format) and a brand-new open-source DLIS writer replacing closed-source legacy tools.
➡️ UX/UI Modernization: A cleaner, more consistent interface designed for efficiency and better error handling.
➡️ GeoSlicer Remote: Improved performance for multi-user support on a single node, enabling better collaboration.
➡️ Quality Assurance: Our test suite grew from 219 tests to 364, achieving a 100% pass rate on unit tests (up from 86.3% last year!).

🤝 A Heartfelt Thank You
This progress wouldn't be possible without the synergy of a brilliant community.
A special thanks to the entire LTrace team for their tireless dedication to the code and the vision.
We also extend our deep gratitude to our partners at Petrobras and Equinor Brasil. Your support and collaboration ensure that GeoSlicer remains at the cutting edge of industry needs, helping to democratize high-end digital rock physics for everyone.

#GeoSlicer #DigitalRockPhysics #OpenSource #LTrace #Petrobras #Equinor #OilAndGas #Innovation #Retrospective2025

LTrace is proud to participate in the “Multiphase flow in porous media for petrophysicists” workshop, hosted by the SPWLA Brazil Chapter in Rio de Janeiro. Our researcher, Rafael Arenhart, joined this intensive week led by Prof. Martin Blunt from Imperial College London to explore pore-scale physics from the classic fluid flow foundations to the latest multiscale modelling developments.

The course covered the complete technical journey, starting from image acquisition to fluid modeling, moving into the complexities of pore networks, and concluding with the generation and interpretation of pressure and permeability curves. These fundamental principles are the backbone of the GeoSlicer simulation modules, and the insights gained are now helping us further refine our existing multiscale simulation tools.

These advancements find great use in critical applications like oil recovery optimization and CO2 storage modeling.

We want to express our sincere gratitude to Prof. Martin Blunt for sharing his immense expertise. Beyond the technical depth, we were truly impressed by how he maintained such high energy and kept the audience fully engaged throughout 40 hours of lecturing in a single week!
Thank you to SPWLA Brazil for organizing this excellent opportunity for technical growth.

#LTrace #Geoslicer #DigitalRock #Petrophysics #EnergyTransition #PoreNetworkModeling #SPWLA #Innovation

Last week, teams from LTrace Geosciences and Equinor gathered in Bergen, Norway, together with our partners at Simula Research Laboratory, for an energizing workshop marking the official start of the new R,D&I ANP project:

Assisted Borehole Image Interpretation with Machine Learning (AMaBILI).
This initiative—executed by LTrace in partnership with Equinor Brazil—aims to push the boundaries of subsurface understanding by combining cutting-edge machine learning with domain expertise in borehole image interpretation.

🔍 AMaBILI will focus on developing ML-assisted tools to enhance efficiency, consistency, and insights in geological interpretation workflows, ultimately supporting smarter decision-making in complex carbonate reservoirs.

The workshop set a strong foundation for the months ahead:
✨ Clear technical alignment
✨ Collaborative roadmap planning
✨ Exciting discussions on innovation opportunities
✨ Excess-k project handover
✨ A shared commitment to delivering impactful research and technology

A huge thank-you to the Equinor and Simula Consulting teams for the productive sessions and enthusiasm. We’re looking forward to driving this project forward together!

#LTrace #Equinor #SimulaConsulting #AMaBILI #MachineLearning #SubsurfaceInnovation #RDI #EnergyInnovation #BrazilNorwayCollaboration

We are pleased to share LTrace's participation in the SBGF Conference 2025. It was a productive experience for sharing our work and connecting with the geophysics community.

LTrace was represented by Leandro Passos de Figueiredo and Elias Sant'Ana 🏳️‍🌈, who presented two research works:
➡️ Leandro Figueiredo presented the work entitled: "AI for seismic data reduction and model updating"
➡️ Elias Santana presented the work entitled: "Bayesian approach for seismic inversion and facies classification using IP and VP/VS parametrization"

It was a great opportunity to present our research in AI and Bayesian inversion and to engage with colleagues in the field.

Thank you to our co-authors Ricardo Zavala-Torres and Bruno Barbosa Rodrigues, DSc, to everyone who attended our presentations, and to the SBGF for hosting the event.

#SBGF2025 #Geophysics #SeismicData #MachineLearning #AI #BayesianInversion #LTrace

Last week, Ingrid Bertin, Ph D. represented LTrace at the American Association of Petroleum Geologists (AAPG) International Conference & Exhibition (ICE) 2025 in Rio de Janeiro.

She presented the work titled "Integration of multi-resolution X-ray microCT images using machine learning and geostatistical approaches." The presentation demonstrated how machine learning and geostatistics can be used to integrate multi-resolution X-ray microCT images of rocks.

The work is part of a R&D project in partnership with Equinor Brazil. Thank you Ingrid Bertin, Ph D. for the talk and for representing us at the Event.

#LTrace #Geoscience #AAPGICE2025 #Innovation #MachineLearning #DigitalRock

Our Bayesian Linearized Inversion (BLI) plugin is now fully updated and compatible with Petrel versions 2022, 2023, and 2024.

Stop the clock on your seismic inversion workflows. The BLI plugin allows you to accelerate reservoir characterization projects without compromising on quality.

Key benefits include:
⚡ Unmatched Speed: Achieve results 6-10x faster than traditional Constrained Sparse Spike Inversion (CSSI) methods, thanks to our analytical, non-iterative solution.
💻 Full GPU Acceleration: Harness the power of your NVIDIA hardware. Our CUDA-enabled process delivers an additional 4-7x performance boost for elastic inversions.
✅ Trusted Scientific Rigor: Generate high-quality, defensible acoustic and elastic property models with results comparable to slower, optimization-based techniques.
Our robust analytical Bayesian framework makes it possible. Spend less time waiting and more time on high-impact interpretation.

Accelerate your workflow today. Find the plugin on the SLB Ocean Store: https://lnkd.in/dStvsQ_q

#Petrel #SeismicInversion #Geophysics #ReservoirCharacterization #OilAndGas

📢LTrace is excited to announce a data visualization update to our open-source digital rock platform, GeoSlicer! 🚀

We have updated the visualization capabilities in GeoSlicer for single-phase absolute permeability simulations in Pore Network Models (PNMs). While the absolute permeability values are the primary output of the simulations, visual analysis of flow data provides essential context for interpreting numerical results.

The new visualization options include:
➡️ Pressure Field Visualization: The capillary pressure of pores, resolved and unresolved, can now be mapped across the network, and the connectivity can be assessed through the visualization of throats.
➡️ Mapping of Resolved and Unresolved Porosity: The pore structure can be visualized with clear identification of resolved pores and unresolved regions.
➡️ Fluid Flow Vectors: It is now possible to render local fluid flow vectors throughout the network. This directly visualizes preferential flow paths, stagnant regions, and flow tortuosity.

A key application of these tools is the analysis of flow in heterogeneous media, the main subject of the multiphase pore network models of GeoSlicer. For instance, the discontinuity of high-magnitude flow vectors in regions with significant unresolved porosity serves as a strong indicator that fluid percolation occurs through pathways with both the resolved unresolved porosity. This is particularly relevant for the characterization of complex systems like carbonates.

#GeoSlicer #DigitalRockPhysics #PoreNetworkModeling #OpenSource #Geoscience #DataVisualization #DigitalRock #ComputationalScience

We are proud to share that Ingrid Bertin, Ph D. represented LTrace Geosciences at the GEOENG Petroleum UNICAMP 2025 conference!

She delivered a short course on GeoSlicer, our open-source AI platform for Digital Rock images, showcasing how technology can accelerate insights in rock physics and reservoir characterization.

A special thanks to Ingrid Bertin, Ph D., SPE Unicamp Student Chapter, Najlah Zeitoum, MSc and Tania Casas. We also thank everyone who joined and participated in the course — your engagement and questions made it a great experience!

#DigitalRock #GeoSlicer #AI #Geosciences #opensource