How it works
StrataSight leverages AI and machine learning to classify facies and lithologies across the subsurface, using seismic and well data in combination. By integrating multiple seismic volumes with facies and lithology logs, StrataSight identifies subtle patterns in seismic response associated with different rock types and depositional environments.
The system generates a volumetric facies and lithology prediction that is calibrated to well data, enabling interpreters to map geological variation across the survey area with confidence.
Key benefits
-
Detailed geological understanding: Visualize facies and lithology distributions away from well control, supporting reservoir characterization, stratigraphic interpretation, and depositional analysis.
-
Reliable predictions: Classification is grounded in well logs and constrained by seismic attributes, providing defensible insights for exploration and development planning.
-
Workflow-ready outputs: Predictions are delivered in standard seismic formats, ready for integration with interpretation platforms and existing geological models.
StrataSight allows geoscience teams to turn seismic and well data into a comprehensive geological view, reducing uncertainty and supporting more informed subsurface decisions.
Data Requirements
Deliverables
StrataSight is designed to work with the datasets geophysicists already have. To get started, we require:
-
All available post-stack seismic data — StrataSight can operate with a single post-stack volume, but benefits from the inclusion of additional stacks such as near, mid, far, angle, or offset stacks where available. The system integrates all provided seismic data within the recorded bandwidth, preserving amplitude, phase, and wavelet behaviour while reducing uncertainty in the predicted facies and lithologies.
-
At least three wells with facies or lithology logs — these wells provide calibration between the seismic response and observed lithologies. Facies and lithology logs anchor the prediction in real subsurface data and allow the model to learn the relationship between seismic response and subsurface composition across the stratigraphic intervals of interest.
-
Geological context for the area — such as known stratigraphy, structural framework, or interpreted horizons. These inputs help guide the classification and ensure results remain interpretable and consistent with the geological setting.
With this dataset, StrataSight generates a facies and lithology prediction volume calibrated to well control and constrained by the seismic response, extending lithological insight across the full seismic survey area while quantifying uncertainty where data support is limited.
StrataSight provides interpreter-ready outputs designed to integrate directly into existing subsurface workflows, extending facies and lithology insight across the full seismic survey area.
-
Facies and lithology prediction cube — A seismic-derived volume predicting facies and lithology across the entire survey. The prediction is calibrated to well data and constrained by the seismic response, allowing interpreters to map lithological variations away from well control. Delivered in standard SEG-Y format, ready for loading into your interpretation platform.
-
Key attribute volumes — A set of the most influential seismic attribute cubes used by the model to generate the facies and lithology predictions. These volumes provide transparency into the drivers of the prediction and offer additional tools for analysing lithological heterogeneity.
-
Uncertainty products — Uncertainty estimates indicate where the seismic data strongly supports the facies/lithology prediction and where confidence is lower, helping interpreters assess prediction reliability and geological risk.
-
Project delivery timeline — Typical StrataSight projects are delivered within a few weeks from project kick-off, depending on dataset size and complexity. This allows teams to incorporate facies and lithology predictions into ongoing interpretation and reservoir evaluation workflows without long processing cycles.