In collaboration with Veracio (formerly Boart Longyear) and their extensive expertise, we developed the TruStructure application — a cutting-edge, web-based platform for structural and geotechnical logging.

Challenge

To optimise exploration strategies and resource development with a streamlined geological core logging procedure.

Approach

In collaboration with Veracio (formerly Boart Longyear) and their extensive expertise, we developed the TruStructure application — a cutting-edge, web-based platform for structural and geotechnical logging. This innovative solution harnesses the power of deep learning segmentation models to analyse drill core imagery acquired through the TruScan system, delivering precise, reliable and unbiased insights into the structure and integrity of core samples.

By employing semantic segmentation, a pixel-level classification technique, the TruStructure application identifies and delineates key features within the images associated with structural logging. These features include the core itself, orientation lines (olines), borders, veins, fractures, and the background. Leveraging a dataset meticulously labelled by geological experts, a custom deep learning model has been trained to autonomously infer and segment these classes.

What sets the TruStructure application apart is its custom model's ability to maintain high-resolution representations throughout the network architecture. Through a strategic fusion of high-resolution feature maps from lower-level layers and lower-resolution feature maps from higher-level layers, the model effectively preserves fine-grained details and captures comprehensive contextual information, ultimately yielding highly accurate segmentation results.

Empowered by this advanced machine learning solution, the TruStructure application empowers geologists and mining professionals with unparalleled precision in analysing the structural and geotechnical characteristics of drill core samples, all whilst eliminating the manual handling of core. By enhancing the understanding of core structures and rock integrity, the application significantly contributes to efficient decision-making in drilling programs, the delineation of mineralisation zones and resource development. With TruStructure, stakeholders can make informed choices, optimise operations, and drive successful mining endeavours.

Outcome

The development of this simple tool provides numerous benefits, including:

 

Accurate and Reliable Analysis: TruStructure provides precise and reliable analysis of structural features in drill core samples, ensuring accurate understanding of core structures and material composition.

Enhanced Decision-Making: By delivering detailed insights into the structure and geotechnical nature of core samples, the TruStructure application supports efficient decision-making in drilling programs, mineralisation zone delineation and resource development - enabling stakeholders to make informed choices and optimise operations.

Streamlined Workflow: With its web-based platform and advanced machine learning capabilities, TruStructure streamlines the structural logging process, eliminating the manual handling of core, saving valuable time and resources. It automates the inference of segmentation classes, eliminating the need for manual analysis and increasing operational efficiency.


Case Studies

We have a strong reputation for taking on complex challenges. Find out more about the ground breaking work and projects we’ve undertaken.

  • All
  • Defence
  • General
  • News
  • Press Releases
  • Enterprise
'}}
Leading Australian defence technology provider Aurizn Group sets the stage for global expansion with acquisition of McR Defence and new leadership
Following the strategic acquisition of McR Defence from the national independent law firm, McCullough Robertson, in August, Brett Sangster has been appointed as the CEO of Aurizn Group, with Andrew Tymms appointed as Board Chair.
'}}
Veracio
In collaboration with Veracio (formerly Boart Longyear) we developed the TruStructure application — a cutting-edge, web-based platform for structural and geotechnical logging.
'}}
Wine Australia
We used Machine Learning and high resolution satellite imagery to scan the Australian continent for the location of vineyards.
'}}
Digital Flare Model
Aurizn developed a physics-based and validated Digital Flare Model for DSTG to enable simulated testing of scenarios to improve platform survivability.
'}}
Active Movement Extent Discrimination Apparatus: Technology Refresh
We ran a technology refresh to the Active Movement Extent Discrimination Apparatus, which is used by NASA to support research into astronaut rehabilitation strategies to counter long term effects of micro-gravity exposure.
'}}
Electro Optic Survivability Integration Laboratory (EOSIL)
We supported DSTG with a technical refresh of its turn-key high-precision test and evaluation capability for Threat Warning Systems to support more efficient survivability assessments.
'}}
Essential Energy – Vegetation Monitoring
By improving risk profiling of vegetation using remote geospatial sensing and AI, we helped power distributors improve electricity supply to millions across their network.
'}}
Cross Domain Desktop Compositor
Working closely with CSIRO, we developed a new cross domain desktop compositor to advance data security for Defence.
'}}
Tactile Cueing System
We accelerated the technology maturity of a Tactile Cueing System from TRL 3 to TRL 7, resulting in improved outcomes for pilots when hovering.
'}}
Electromagnetic Spectrum Synthetic Environment
We supported continuous evaluation of Electronic Warfare Self Protection (EWSP) systems and Tactics, Techniques and Procedures (TTPs) to improve effectiveness.
'}}
Tactics Development Experimentation
Through the AWAKENING Series of Experiments, Aurizn provided the ADF with insights into improving aircraft capabilities and platform survivability against land-borne ballistic weapons.
'}}
IGO
Using deep learning, we worked with miner, IGO, to revolutionise the minerals exploration process via AI-generated prospectivity heatmaps.