You are Home   »   News   »   View Article

Geoteric’s neural networks for subsurface

Wednesday, May 3, 2023



Seismic interpretation software company Geoteric is offering two new 'neutral network' based tools for understanding the subsurface, and now offers six such tools in total. They will be added to Geoteric's 'AI Faults - 3D networks' module, in a new release called Geoteric 2022.2.

The company says its AI Faults technology package can reduce the time needed to interpret faults by up to 95 per cent. It is used by 40 companies.

The software can be used to extract fault surfaces into other modelling packages, such as software for building geocellular models.

The 'Acorn' tool can be used to train a neural network from scratch and then fine-tune it, on a company's own seismic data, or basin, reservoir or field data.

Users can choose between different pre-trained networks or use an untrained network. By starting with an untrained network, they can keep the entire ownership of the subsequent trained network within the company.

Users can share neural networks between projects and between colleagues and teams, but keep the company's intellectual property within the company.

The 'Meranti' tool is a pre-trained network. Users can fine-tune their models, using 'fault sticks', as they do on more traditional interpretation systems.

'Some users prefer to use the software straight out of the box to get instant results, knowing the time, effort and investment of training a [neural] network to a high standard using multiple data has already been completed,' says Mark Brownless, CTO at Geoteric.

'Other users want to have more control over their data and results. They want to be able to choose how they educate the network to a particular style of geology or seismic data.

'We are empowering them with options so they can decide what is right for them and their project.'



Associated Companies
» GeoTeric

CREATE A FREE MEMBERSHIP

To attend our free events, receive our newsletter, and receive the free colour Digital Energy Journal.



DIGITAL ENERGY JOURNAL

Latest Edition Mar-Apr 2025
Apr 2025

Download latest and back issues

COMPANIES SUPPORTING ONE OR MORE DIGITAL ENERGY JOURNAL EVENTS INCLUDE

Learn more about supporting Digital Energy Journal