Tela™ agentic-AI assistant for geological analog search

Return to Tela catalog

Screen massive 3D seismic volumes for high-potential prospects in a fraction of the time.

Challenges

• Identifying prospective oil and gas traps within large 3D seismic volumes can be time-consuming.
• IGeoscientists often rely on finding geological analogs with characteristics similar to proven prospects.
• IScreening seismic datasets manually can take weeks or months of interpretation effort.
• ILarge data volumes make it difficult to consistently evaluate all potential opportunities.
• IManual analog selection can introduce bias and variability into prospect evaluation.

Solution

• Leverage a pretrained seismic foundation model that understands seismic patterns and geological features.
• Accept either a proxy model of a successful analog or an example prospect as a reference target.
• Search across large seismic volumes to identify similar geological features automatically.
• Generate probability-ranked analog candidates for further interpretation and evaluation.
• Deliver results in minutes rather than weeks, accelerating exploration workflows.

Features

• Cross-Survey Heatmap Generation – Generate probability heatmaps across multiple seismic surveys simultaneously, enabling direct comparison of analog quality across datasets. Color-coded visualizations help geoscientists quickly identify areas with the highest similarity to the reference target.
• Ranked Analog Results with Geographic Context – Return probability-ranked analog candidates in a searchable table and geographic map view. Users can evaluate similarity scores, locations, and spatial context to rapidly prioritize exploration opportunities.
• Petrel Integration and 3D Visualization – Import identified analogs and heatmaps directly into Petrel for 3D visualization and interpretation. Move seamlessly from regional screening to detailed structural and stratigraphic analysis within existing workflows.

Benefits

• Accelerate Exploration Screening – Reduce analog hunting from weeks or months to minutes using AI-driven similarity analysis.
• Improve Consistency and Objectivity – Apply repeatable, model-driven analog identification across large datasets.
• Reduce Exploration Uncertainty – Support prospect evaluation through consistent analog ranking and comparison.
• Scale Across Large Seismic Volumes – Screen significantly larger datasets than would be practical through manual interpretation.
• Integrate Seamlessly into Existing Workflows – Leverage direct Petrel integration for rapid transition from screening to interpretation.