AI in daily mining operations

[Tovael]

I've noticed AI making its way into daily mining operations, especially for predictive maintenance and refining workflows. However, not every model seems to deliver as promised.

I'm curious what successes have you had, what didn't pan out, and where do you think machine learning is genuinely beneficial?

 

[rockbreaker7]

After implementing an AI monitoring system we were able to identify pump failures weeks in advance, which was a game changer. On the flip side, its predictions for ore grades were inconsistent. It seems to perform best when the data is clear and consistent, but it struggles when the geology gets complicated

 

[Safari]

Our team tried out machine learning to optimize the plant and set up predictive alerts.
It helped us plan for downtime better, but it didn’t quite deliver on real-time solutions.

 

[Zivon]

We gave AI a shot for haul truck maintenance since we were tired of unexpected breakdowns. It did a great job at spotting bearing failures early on, but it had a hard time with issues caused by operators. While it excels at analyzing trends and reading sensors, it can’t quite replace the experience and intuition of seasoned workers underground.