Speaker: Rocky Roden
Summary
Artificial Intelligence (AI) and Machine Learning (ML) are redefining geoscience interpretation by enhancing workflows rooted in centuries-old physical laws. These technologies are already solving challenges like stratigraphic analysis, fault detection, and facies distribution, while future innovations—such as digital twinning, automated processes, and synthetic modeling—promise even greater transformation. Large Language Models (LLMs) bring new possibilities for interacting with geological data, offering advanced text summarization, translation, and coding support.
Seismic processing is also advancing through Full Waveform Inversion (FWI), delivering high-resolution insights that may revolutionize seismic interpretation. Supported by today’s HPC and cloud infrastructure, the horizon includes quantum computing, which could exponentially accelerate computational capabilities within a decade.
Discover how these groundbreaking technologies are reshaping geoscience workflows. Register now for this exciting webinar by clicking the button on the right of the page!
Detail
Every day our lives are intertwined with applications, services, orders, products, research, and objects that are incorporated, produced, or effected in some way by Artificial Intelligence and Machine Learning. However, it is not always clear how AI/ML solutions relate to interpreting and understanding geology. Geoscience interpretation is grounded in physical laws discovered by the scientific community over multiple centuries. Machine learning is a data driven solution that provides an augmented or different approach to interpretation workflows. In geoscience interpretation, there have been numerous AI/ML approaches developed over the last few years to solve specific singular interpretation tasks such as stratigraphic analysis, fault and fracture detection, and facies distribution. Future trends include the expanded application of synthetic models and digital twinning, automation of interpretation processes, and the combining of machine learning approaches. Recently Large Language Models (LLMs) have evolved that are designed to understand and interpret human language based on massive amounts of data. LLMs infer from context, generate coherent and contextually relevant responses, translate to languages, summarize text, answer questions and assist in writing or code generation tasks. Will LLMs be the future of interacting with our geological and geophysical data?
Another significant geoscience technology trend that incorporates several machine learning approaches is Full Waveform Inversion (FWI) in seismic processing. There have been some dramatic results over the last few years where FWI iteratively updates an estimated subsurface model and computes corresponding synthetic data to reduce the difference (the data misfit) between the synthetic and recorded data. The objective of FWI is to match the synthetic and recorded data in a comprehensive way, such that all information in waveforms (e.g., travel times, amplitudes, converted waves, multiples, etc.) is accounted for in the data misfit. Today, FWI is primarily employed as an effective tool for high-resolution velocity model building (acoustic FWI). However, recent elastic FWI results have suggested that this technology may change the way we process and interpret seismic data in the future.
Today the necessary computer power and associated architecture for advanced technologies like AI/ML and FWI are basically centered around CPU/GPU configurations, High Performance Computing (HPC), and the Cloud. However, most experts indicate within a decade, quantum computers will be fundamentally operational and estimates of the increases in today’s computational speed could be from millions to exponentially faster. What would a world with quantum computers be like?