A recent research paper from CSIRO, Australia, introduces ML-SEISMIC, an innovative solution that tackles the challenges of traditional methods used in geological investigations. ML-SEISMIC is a physics-informed deep neural network designed to autonomously align stress orientation data with an elastic model.
ML-SEISMIC’s unique approach eliminates the need for meticulous manual adjustments of geomechanical properties and boundary conditions, providing a more streamlined and powerful process for stress and displacement field estimations. Its methodology employs physics-informed neural networks to solve linear elastic solid mechanics equations, optimizing stress field eigenvalues for comprehensive understanding of stress and displacement fields.
The application of ML-SEISMIC in Australia showcases its ability to autonomously retrieve displacement patterns, stress tensors, and material properties. Notably, the network utilizes Global Navigation Satellite Systems (GNSS) observations to revisit large-scale averaged stress orientations and identify areas of inconsistency, highlighting its adaptability across various scales.
In conclusion, the groundbreaking ML-SEISMIC emerges as a transformative solution in geological investigations, offering a reliable interpolation framework and promising insights into Earth’s dynamic processes.
Check out the full paper for details by the researchers in this project here.
And, don’t forget to join our ML SubReddit, Facebook Community, Discord Channel, LinkedIn Group, and Email Newsletter, where we share the latest AI research news, cool AI projects, and more.
If you like our work, you will love our newsletter. Subscribe to our newsletter here.