Bentley Systems mining and energy software business Seequent says its new AI-driven Imago image analysis tool enables geologists to analyse drill core imagery in minutes rather than days or weeks.
“Every core tray represents a substantial investment yet core logging has traditionally been a manual, qualitative process with results influenced by the training, experience, time and resources available to the geologist or core logger,” Seequent drilling data product line manager, Tim Schurr, said.
“By applying AI within Imago’s established workflow, Imago image analysis brings repeatability to the process, helping organisations save time and money while improving the consistency of the data captured.
“That matters for every decision that follows.
“Geologists can investigate unexpected results, challenge assumptions and validate and strengthen the models that guide drilling decisions.”
Bentley acquired Seequent in 2021 and added Imago’s cloud-based software for the capture and management of geoscientific imagery soon after.
Schurr said AI embedded directly into existing exploration workflows could improve the consistency of and confidence in geological data that underpinned exploration decisions instead of simply accelerating tasks.
Imago image analysis used Seequent’s proprietary machine-learning models to transform core photo libraries collected across years of drilling programs into quantitative geological data. Geologists could review outputs against underlying core imagery, adjust analyses and validate results before using them in modelling, interpretation and exploration decision-making – “while preserving a direct connection to the rock itself”.
The connection extended the value of core imagery and geological data beyond initial core logging, supporting interpretation and decision-making throughout the exploration and mining lifecycle.
Purpose-built machine-learning models were being used to automate core image preparation and analysis, converting new and historical drill core tray images into outputs such as downhole datasets for Rock Quality Designation (RQD).
Seequent cited an example where 600m of core imagery had been processed using Imago image analysis in two minutes to produce “structured, depth-referenced drilling data for downstream interpretation”.
