KEY POINTS
- BME is using AI, drones and 3D modelling to improve mining blast decisions.
- A Gauteng quarry trial cut D50 fragmentation by 8.45% and overall measured fragment size by 11.21%.
- The company says AI will support mining experts rather than replace human expertise.
BME, a division of Omnia Holdings, is using artificial intelligence (AI), advanced data analytics and digital technologies to improve blasting operations and help mining companies extract greater value from every blast.
The company says the approach is designed to turn the large volumes of operational data generated throughout the mining process into practical insights that can improve decision-making, safety, productivity and downstream processing.
The technology was discussed during BME’s webinar, “Reimagining Blasting: Unlocking Digital Mining Value Through AI-Enabled Solutions”, where company executives explained how AI can be applied across the mining value chain without replacing the expertise of engineers, blasting specialists, geologists and other professionals.
Nishen Hariparsad, BME’s General Manager for Technology and Marketing, said the company is focused on using operational data to generate insights that can support better decisions throughout the blasting process.
Rather than viewing AI as a replacement for human expertise, BME considers it a tool that can give mining professionals better and more reliable information.
Hariparsad described the company’s approach as “integrity-enabled AI”, which is based on trusted and validated operational data.
According to him, BME’s AI solutions are designed to be transparent in how data is generated, consistent with safe and responsible mining practices and supportive of human decision-making.
He stressed that future mines would continue to require experienced professionals, including engineers, geologists, blasting specialists, operators and other technical personnel.
Christiaan Liebenberg, BME’s Product Manager for Software, said the company has developed a vertical AI model specifically for the mining industry.
The model incorporates a broad range of mining data, including geological conditions, drilling information, blast designs, charging practices and fragmentation results.
BME is also developing a mine-to-mill approach that combines digital twins and machine vision to create a virtual representation of the mining environment.
This technology allows the characteristics of rock to be tracked from the initial blast through the crusher and into the processing plant.
The objective is to help mining professionals identify potential downstream consequences before a blast takes place, rather than only discovering problems after material has reached the processing stage.
AI Learns Across Multiple Mining Operations
One of the key features of BME’s system is its ability to learn from data generated across different mining operations.
Through federated learning, information from individual sites can contribute to improving the wider AI model without exposing confidential customer information.
BME said site-specific data is anonymised and stripped of sensitive information before only the aggregated information required to improve the central model is incorporated.
This means individual mining operations can contribute to a growing body of knowledge while keeping their proprietary datasets within authorised project environments.
As more blasts are analysed, the system can identify recurring patterns and provide increasingly useful guidance to blast engineers, supervisors, shotfirers and foremen.
The company believes this could allow knowledge that was previously confined to individual specialists or individual mine sites to become part of a broader, continuously improving knowledge base.
BME has invested in Xplosmart® as the AI platform supporting its digital blasting strategy.
The integration of the technology is being implemented in phases, beginning with BlastMap®, which is currently under development.
The company plans to follow this with Xplolog®, a data-capture system designed to collect information such as drilling depth, charging and stemming values and feed the data back into the wider system.
The combination of these technologies is intended to create a more complete digital record of each blast and provide better information for subsequent decision-making.
BME demonstrated the practical application of the technology during a trial at a hard-rock granite quarry in Gauteng, South Africa.
Hendrik Hougaard, BME’s Senior Blast Technician, said the quarry had been experiencing excessive oversize material following blasting.
Large rock fragments were affecting the mine-to-mill process by reducing operational efficiency, increasing fuel consumption and equipment wear, while also creating additional costs associated with secondary breaking.
The geological conditions made the problem more complicated, with significant fractures, joints and other discontinuities present across the rock face.
BME surveyed two blast blocks to establish a baseline and then determine whether a targeted change to the blast design could improve fragmentation. Using drone-based photogrammetry, BME created a three-dimensional model of the first blast area.
The survey itself took about 10 minutes, while generating the 3D model took approximately 90 minutes.
The model provided a more comprehensive view of the rock face than conventional two-dimensional profiling, allowing the company to assess burden conditions across the entire face and in front of individual drill holes.
The analysis showed that the blast geometry, including burden, spacing and drilling compliance, was generally acceptable. However, it identified areas where burden was either excessive or insufficient.
One of the findings was a potential corner failure that could have increased the risk of face burst, flyrock and elevated airblast levels.
The analysis also identified a significant underburden section behind the potential failure area, allowing adjustments to be made to charging before the blast was conducted.
Following the initial assessment, BME focused on improving fragmentation by changing the timing between blast rows.
The first blast recorded a D50 fragmentation size of 509 millimetres and a D18 of 1,011 millimetres.
For a quarry supplying a jaw crusher with a specified maximum feed size, the company said excessive oversize can have implications for loading, hauling and downstream processing.
BME subsequently shortened the inter-row timing by 25 milliseconds to improve interaction between blast rows and enhance the crushing and heaving action of the muckpile.
The second blast block was again modelled in 3D, while another drilling-compliance audit was carried out.The rock mass was also analysed using RockMass AI, with additional attention given to burden profiles.
The results showed an immediate visual improvement, with fewer large fragments visible at the top of the muckpile.
Post-blast analysis using Xplosmart®’s fragmentation AI module showed that the D50 fell from 509 millimetres to 466 millimetres.
The D18 also declined from 1,011 millimetres to 984 millimetres.
BME said the changes represented an 8.45% improvement in D50, a 2.64% improvement in D18 and an overall 11.21% reduction in measured fragment size.
Hougaard described the change in D50 as operationally significant for a quarry, noting that even a relatively small improvement at the blasting stage can produce larger benefits further down the production chain.
The company plans to continue testing other variables at the quarry, including burden variability, initiation practices, blast patterns, burden and spacing.
These changes will be introduced incrementally while maintaining the powder factor and overall blast costs, allowing the impact of each adjustment to be measured.
Despite its growing investment in AI and automation, BME says people will remain at the centre of its digital transformation strategy.
Hariparsad said technology alone does not create value, arguing that value comes from professionals using better information to make informed decisions.
The company’s strategy therefore focuses on combining AI-generated insights with the knowledge and judgement of experienced mining personnel.
BME believes this approach could help mines improve fragmentation, reduce waste and secondary breaking, enhance safety and increase efficiency across the mine-to-mill process.
The company’s broader objective is to enable mining professionals to extract more value from every blast while maintaining the central role of human expertise in the industry’s increasingly digital operating environment.
