Shearwater has outlined how its Reveal ML suite integrates machine learning directly into established seismic-processing and imaging workflows, moving the technology beyond isolated, task-specific applications.
The tools address production challenges including complex deghosting, data-driven denoising, swell-noise removal, adaptive subtraction, segmentation and post-imaging processing. They can be applied to both streamer and ocean-bottom node data without requiring processors to write code.
A built-in training environment also allows users to develop models for their own datasets and deploy them at scale. The underlying framework records experiments, metrics and trained models, supporting reproducibility and comparison rather than treating machine-learning results as one-off black-box outputs.
Additional functionality and a software-development kit are planned as part of the Reveal 8 release in November 2026, enabling users to develop or integrate their own machine-learning tools.
