top of page
ESA_WaveVertical.png

EARTHAI

Propagate knowledge
from core to seismic scale

EarthAI grey.png

EarthAI is where human expertise and Machine Learning come together to generate outputs from your data. 

The module allows inputs to train and apply Machine Learning models to well data from wireline logs, core data, fluid data and CPI’s, as well as seismic data. It also allows well data to be propagated to the seismic scale and 3D property cubes to be generated in a matter of months. When working together, EarthAI and expert knowledge facilitates log data imputation and petrophysics property predictions. 

 

Don’t lose out on valuable insights as a result of gaps in data or limits on time – EarthAI boosts labelling efficiency and utilises Machine Learning to identify new data relationships and de-risk your subsurface studies. 

Asset 40.png

USES

EARTHAI WELLS


Rapidly predict high-quality rock and fluid property curves and predict missing logs in wells

With EarthAI Wells, geoscientists can create and use Machine Learning models to predict reservoir properties directly from wireline logs, core data, fluid data and CPI’s.

 
By making use of Machine Learning methods for property prediction, geoscientists can carry out projects at a large scale efficiently. By freeing them from cumbersome, time-consuming activities, they can spend more time on high value activities, scaling their expertise from prospect level accuracy to basin level coverage.


An example of this is the “missed pay project”, where pay intervals are predicted and characterised for more than 8000 wells in the Northern North Sea. Read more here.

Robust data-driven methods accelerate 1D well based workflows

 

  • Fast, accurate, unbiased and repeatable workflows
     

  • Predict missing curves and properties using AI and expert inputs
     

  • High quality model predicted data increases understanding of uncertainty 
     

  • Petrophysics and well log interpretation type tasks in minutes not days
     

  • Perform rock and fluid characterisations with state-of-the-art viewers 

EarthNET log view.png

EARTHAI INTERPRETATION

Dramatically increase the efficiency of Seismic Interpretation workflows with Automatic Seismic Interpretation

Reduce your workflows from weeks and months to days, while simultaneously improving the quality of interpretation.  By easily training and/or applying Machine Learning models for automatic seismic interpretation, geoscientists can produce repeatable and highly detailed structural and stratigraphic interpretations with tools to analyse associated uncertainties.


These models can be further refined to include local geological knowledge, which can be incorporated into global models and vice versa, thereby enhancing your interpretation activities.

EARTHAI_Interp.png
ESA_Wave.png
ESA_Arrow.png
  • Take control of your seismic with the support of pretrained models and manual assisted labelling
     

  • Exceed the capabilities of traditional seismic workflows with AI-assisted fault interpretation, horizon interpretation, geobody interpretation and litho-seismic interpretation
     

  • Propagate knowledge from the well to seismic scale to predict reservoir properties from elastic properties generated from well data, as a function of 3D partial stacks, or as a function of partial-stack cubes
     

  • Use well-ties to ensure precise feature-to-label alignment

esa-image-2-.png

EARTHAI VOLUMETRIC


Rapidly predict high-quality rock and fluid property cubes
 

With the latest addition to the EarthNET toolkit,  3D rock-fluid property predictions can be generated either as a function of inverted seismic data, or as a function of partial-stack cubes.  


The challenge of inversion is that it is  highly specialised, time consuming and still delivers non-unique results.  With AI, property prediction provides an excellent alternative method to augment the results derived from traditional inversion techniques.

EARTHAI_Vol.png
  • Use regional well data to predict missing log data and tie wells with seismic

  • Models can be trained to predict any desired property

  • No need for data conditioning 

  • Fully scalable, e.g. NNS regional reservoir property prediction model

ESA-new-image

RESOURCES

NEWS

Earth Science Analytics announce collaboration with Horisont Energi

BLOG

Bringing Norwegian Innovation to Houston

EDITORIAL

Cutting down risk of oil and gas exploration by releasing potential in dormant data.

Hart Energy,  January 2022

bottom of page