Ltrace Geosciences

LTrace is a startup formed by physicists and computer scientists with expertise in developing solutions for stochastic seismic inversion, deep learning for Digital Rock Analysis and machine learning applied to reservoir history matching. Seismic inversion is a powerful tool for predicting reservoir properties. Our methodology is based on Bayesian methods, providing a rich analysis of the uncertainty in the reservoir characterization process, helping to reduce the risks of well drilling and maximize oil production. In a R&D partnership with Petrobras, LTrace is developing new Deep Learning techniques for segmentation and identify pretophysical properties and lithofacies from CT and micro-CT. We also work with computer vision applied to 4D seismic integration into reservoir history matching with the ESMDA algorithm. See more about R&D projects About our products We are experienced in digital rocks where we are developing the open source platform GeoSlicer. Developed in a cooperation agreement between LTrace, Petrobras and Equinor, the software consists of an integrated computational platform for visualizing and processing digital rocks in all scales. It is based on the opensource platform called 3DSlicer. Includes several tools for processing and orienting multiple cores, extraction and visualization of profiles, 3D textural segmentation, pore/grain network analysis and upscaling of plug properties to the well log scale. GeoSlicer aims to meet the demands of geoscientists from different areas and expertises, from petrophysics to geological interpretation. Within the context of digital transformation, it implements different methodologies with machine learning bringing data-science to geoscience and it allows the remote access to digitized rocks facilitating working remotely from home-offices or anywhere else.