This blog is the second part of the four-part blog series on using machine learning to reduce cost in drill and blasting. (For part one, please see Using Machine Learning to Make Blasting Design Cheaper, Faster, and More Intelligent.)In this post, we focus on how machine learning can help fragmentation analysis and prediction.
Fragmentation analysis and prediction are useful for quarry and mine operations to improve blast results and reduce cost. Better fragmentation early on will have beneficial knock-on effects on hauling and cleaning cost. Since these stages are constantly repeated, improving blast data and reducing cost will have continuous benefits in both the short and long terms.
Illustration by Scott G. Giltner at Quarry Academy
Fragmentation analysis and prediction is often challenging because the blaster faces a tricky trade-off between fragmentation and swell.
Traditionally, fragmentation analysis and prediction heavily rely on the ground personnel’s experience and trial-and-error. They choose from available fragmentation models to predict the initial blast. Based on the initial result, they iteratively tweak parameters for a more desirable balance between fragmentation and swell. Sometimes they just stick to the initial model and rinse and repeat.
Strayos introduces machine learning to fragmentation analysis and prediction. Our platform automatically learns from the result of each blast and analyzes how drill pattern, timing, and powder factor each contributes to the resulting fragmentation and swell.
With the help of machine learning, the ground personnel can see their blasting improvements more visually and tweak parameters more efficiently and precisely. Quarries and mines can more precisely control cost and predict revenue.
Get the latest posts delivered right to your inbox