Dr. Lingchen Zhu and Dr. Peter Tilke, schientists at Schlumberger-Doll Research, present "Accelerating Geoscientific AI with Synthetic Data" at the MIT Earth Resources Laboratory as part of the Friday Informal Seminar Hoiur.

"Data is crucial for implementing, training and applying AIML solutions for subsurface interpretation. Synthetic data can generate labeled realistic data efficiently and at scale for AIML applications, is non-identifiable so privacy and data security concerns don't apply, can be used to train and test AIML models, for exploratory data analysis, to validate assumptions, demonstrate results that can be obtained with AIML models, and reject models producing poor results without the cost of acquiring and incorporating real data, and can be used for transfer learning.

In this session we would like to share our experiences with generating synthetic geoscientific data (specifically stratigraphic forward models) for the purpose of accelerating AIML workflows. Building a synthetic data library of geologic analogs represents a level of complexity which involves creating, managing and accessing vast amounts of data efficiently for a variety of use cases."

Lingchen Zhu is a Senior Research Scientist with Schlumberger-Doll Research after he graduates his Ph.D. at Georgia Institute of Technology in 2016. After joining SDR, his project portfolio includes but not limited to deep supervised learning on acoustic processing, deep unsupervised learning for signal pattern recognition, probabilistic deep learning to evaluate inference uncertainties, automated machine learning, generative adversarial networks for rapid data generation and conditioning, etc. His current research interest includes building large-scale data library via automated High-Performance Computing pipelines on hybrid clouds and implementing advanced AI/ML systems for automated subsurface modeling and interpretation.

Peter Tilke is a Scientific Advisor with Schlumberger-Doll Research with over 35 years of experience. He has a Ph.D. in Geology from MIT EAPS and previously worked as a research geologist with Shell. At SLB, Peter is a computational geoscientist and has developed many components for SLB's interpretation and modeling products. His current research focuses on AI, and ML approaches to automate and accelerate subsurface interpretation and modeling.

MIT Earth Resources Laboratory's Friday Informal Seminar features guest speakers from industry and academia on topics relevant to our lab, including geophysics, seismology, rock physics, imaging, inversion, machine learning, and the energy industry. Titles and abstracts will be posted here when available. Contact fish_seminar_organizers@mit.edu for more information and Zoom password.