Geophysical Insights hosting the 2018 OIl & Gas Machine Learning Symposium in Houston on September 27, 2018
Dr. Tom Smith presenting on Machine Learning at the 3D Seismic Symposium on March 6th in Denver
What is the "holy grail" of Machine Learning in seismic interpretation? by Dr. Tom Smith, GSH Luncheon 2018
Using Attributes to Interpret the Environment of Deposition - A Video Course. Taught by Kurt Marfurt, Rocky Roden, and ChingWen Chen
Dr. Kurt Marfurt and Dr. Tom Smith featured in the July edition of AOGR on Machine Learning and Multi-Attribute Analysis

Geophysical Insights Launches Paradise 3.2, adding Geobody Identification Through Machine Learning and Attribute Generation using High-Performance Computing 

Houston, 16 July 2018

Geophysical Insights announces the launch of Paradise 3.2, which includes the isolation of geobodies through machine learning, and the generation of attributes on a LINUX High-Performance Computing (HPC) facility. Paradise is a multi-attribute seismic analysis workbench that uses machine learning to extract more information from both seismic and well data.

The new geobody analysis ThoughtFlow™ in Paradise derives geobodies from the Self-Organizing Map (SOM) machine learning process. This new technology in Paradise is the first instance in the industry to generate geobodies automatically using machine-learning classification results. The new geobody capability allows interpreters to isolate areas of interest based on a common neuron or set of neurons. Upon generation, interpreters can edit - prune or fill - geobodies at the individual seismic sample level. Given velocities alone, the geobody analysis tool will estimate volumetrics.  Geobodies can also be filtered for size and neural classes.

Along with geobodies, the newest version of Paradise enables companies that have an HPC cluster to generate AASPI geometric and spectral decomposition attributes in a fraction of the time it would take on even a large, multi-core server. The HPC capability also supports both SLURM and LSF Scheduler formats to ensure resources are managed effectively.  Paradise 3.2 offers other performance improvements, including faster SEGY data loading and a connector to Petrel 2017. 

“The goal of both these new tools, and of Paradise itself, is to increase the efficiency of the interpreter and leverage machine learning technology," states Dr. Tom Smith, President/CEO of Geophysical Insights. “The ability to automatically generate geobodies using machine learning takes the industry to a new level of interpretation. And, including HPC capabilities in Paradise was a logical extension of the Paradise platform due to the increasing use of geometric and spectral decomposition attributes in large seismic volumes. Our goal is to enable every interpreter to leverage machine learning technology in ways that will provide more time for analysis and focus on what’s important - finding oil and gas. We are delighted to offer a dramatic new Paradise version that applies new machine learning processes and reduces the time to results.”

About Geophysical Insights

Geophysical Insights was founded in 2008 by Dr. Tom Smith with the vision of applying machine learning methods to geologic interpretation to reduce the risk of exploration and the cost of field development.  The mission of the business is to develop the next generation of interpretation tools in an intuitive, elegant interface that can be used by all interpreters.  Over three years in development, Paradise commercially launched at the Society of Exploration Geophysicists (SEG) Annual Convention in 2013 and has seen steady adoption in the industry since its launch. 

Geophysical Insights continues to build machine learning capabilities in the Paradise workbench, which is proving to reveal deeper insights into the seismic response. Built for large sets of data, Paradise is a 'Big Data' solution that easily scales from a single workstation to an enterprise. With adaptive algorithms that learn from the data that are enabled by high-performance computing, interpreters are able to extract more information from seismic and well data than would otherwise be possible from traditional interpretation tools.