Simple applications of machine learning in subsurface characterization

Dr. Misra will present case studies on the use of machine learning techniques. In the first case study, neural network models generate NMR T2 distribution in the absence of an NMR logging tool. In the second case study, simple data-driven models generate compressional and shear travel time logs in the absence of a sonic logging tool. In the third case study, machine learning assisted the segmentation of SEM images of shale samples.

This segmentation method involves two steps, feature extraction from SEM images followed by random forest classification of each pixel in the SEM image. In the fourth case study, machine learning was used to process CT scan images to predict the subsurface geomechanical properties.

Speaker Bio

Dr. Siddharth Misra

Dr. Misra holds a Ph.D. in petroleum engineering from The University of Texas at Austin. Prior to that, from 2007 to 2010, he worked as a wireline field engineer in Saudi Arabia, Egypt, and USA with Halliburton. He received his undergraduate degree in electrical engineering from Indian Institute of Technology Bombay, India, in 2007. Recently, he was awarded the prestigious Department of Energy Early Career Award, American Chemical Society New Investigator Award, and SPE Mid-Continent Formation Evaluation Award. His research interest includes subsurface characterization, machine learning, sensing and sensors, and inverse problems.