Skip to main content
search

Artificial Intelligence (AI) is here, and we are all experiencing its impact, fast – from maps and navigation to facial recognition, text editors, human-indistinguishable generated videos and photos, social media algorithms and more.

It’s taking over organisations on a global scale, too, with research from IBM finding more than 40 per cent of enterprise-scale businesses have actively deployed AI within their operations. And despite concerns around some aspects of the technology, such as the potential to drive misinformation online, the opportunities ahead are almost beyond current comprehension. McKinsey’s report on The economic potential of generative AI estimates half of today’s work activities could be automated by 2060, roughly a decade earlier than in previous estimates.

At Seequent, we see a huge opportunity in the value and innovation that can be driven by AI. If used properly, AI can be a major contributor in solving the big problems facing the planet right now – like building more resilient infrastructure, efficient sourcing of critical minerals and so on. However, it comes with its own set of difficulties.

Exploring uncharted depths

Witnessing this first-hand, Dr Alexandre Boucher, Vice President of Seequent Labs and one of the world’s leading geo-statisticians, identifies the vast quantities and unorganised proliferation of data as one of the biggest hurdles to deploying AI for the subsurface. “If we look towards any of the AI success stories – Chat GPT or Microsoft Copilot – they have the capability to structure and then train large language machine learning models using vast amounts of data. But in the subsurface we are dealing with unknowns because the data often isn’t there or it is unstructured; we rely on techniques like mapping and drilling to find it. And even when we do have data, it is often siloed – in project files, or on an individual’s hard drive, held in different file formats and organised in a wide array of different data standards – meaning it cannot easily be combined.”

Machine learning in Earth’s shadows

The world is consuming resources at a faster rate than ever before, demanding geoscientists to find new ways to understand the underground; and do it better, smarter, and with less impact. There is no doubt AI will speed up data analysis, improve workflows and reduce manual errors for the geoprofessionals we work with. It will free up these operationally-critical teams to focus on more meaningful pursuits and deliver business efficiencies and better outcomes.

Several years ago, we identified that to help our customers use AI effectively within processes and workflows, we needed to dismantle the barriers limiting interoperability and liberating the data. This starts with a platform that seamlessly integrates data of any type or format, unlocking AI’s full potential for subsurface applications.

It all starts with the data: our data-centric ecosystem

That’s why we’re building a unified and open data platform: The platform  will be open and our customers can easily integrate solutions and data sources, whether from Seequent or any other provider. It is designed to meet the rigorous data security and privacy requirements of the world’s largest organisations.

Seequent is investing in the future of our planet by building this unique platform that will allow companies to build the deepest understanding of the underground that has ever been seen, by removing barriers to understanding. It is the connected future of technology for the subsurface. We can see it disrupting and transforming industries, at scale, in the years to come says Dr Boucher.

A leading player in the mining industry, and one of Seequent’s customers, has been a part of the early access program for the new data platform and sees it as a way to expedite their operations and enhance efficiencies, commenting: “We believe in the vision Seequent has, which is why we got involved in the early access program. If there is potential to make better use of our data, and save time in the process, that’s something we should all get on board with.”

Enhancing product capabilities 

Work is also underway to bring in-product AI and Machine Learning (ML) capabilities to our customers’ workflows to automate repetitive tasks and reduce time spent on manual processing, while also speeding up analysis. This will enable geoprofessionals to address higher-value work and improve project outcomes.

ML techniques use AI to enable a computer to learn and develop without explicit instruction. ML has great potential to speed up geoscience tasks, for example the lengthy interpretation of large amounts of field data.

A high-resolution image of subsurface core samples in Seequent’s Imago. New machine learning capability is speeding up the process of logging core image data.

Imago AutoCrop – a nifty new tool in our image processing solution Imago – uses machine learning technology to automatically crop and linearise imagery of samples of subsurface materials, called cores. These core samples drilled from the ground are essential for pinpointing the location of mineral deposits for new and existing mines. As these samples degrade over time, capturing information when they have been freshly drilled is vital. Imago was developed to streamline the laborious task of core logging, which, according to a satisfied AutoCrop customer, has now transformed from hours to minutes.

We’re further enhancing this efficiency with new machine learning-assisted logging features and also running early access programs on cutting-edge data science techniques and the detection of geological trends, integrating AI and ML capabilities across our product portfolio.

Customer-led innovation 

And it’s not just what we’re doing. Our journey is fuelled by our customers’ creativity with our software. The geoscience team at Australian mining and exploration company IGO are building a virtual reality platform to find exploration targets using Seequent’s 3D modelling solution Leapfrog Geo.