Getting value from Digital in Mining: Big data

  • Home
  • /
  • Digital Transformation Blog

Digital Transformation Blog

Practical insights on the digital revolution

Getting value from Digital in Mining: Big data

Around the world, we see lots of hype and interest in applying digital technology to mining operations. We work with many clients to help them identify opportunities to extract value from new technology. Based on what we’re hearing and seeing, we have identified a few key areas where technology can change the game in the short term.


The dramatic advances in democratizing quantitative analysis goes by many names – Big Data, Artificial Intelligence, Machine Learning, Advanced Predictive Analytics, et al. But regardless of the nomenclature, taking a rigorous look at a lot of good data has the potential to deliver enhanced insight into value-adding behaviors.

Many mining companies are piloting Big Data across multiple applications. We have seen robust examples of Big Data technologies applied to:

  • Exploration – getting greater insight from geographical data, even data collected years ago
  • Equipment maintenance – predicting machine failures, identifying behaviors that lead to longer equipment life
  • Improving plant operations – putting artificial intelligence on top of the current process control systems and removing the black art of running a plant. However, this comes with a large caveat-- AI is not a cure-all. It depends on an operation being crisply run already, with strong planned maintenance fundamentals. Without that firm foundation, Big Data on predictive maintenance value is limited as you are just building a bigger backlog or increasing non-compliance.

Most of these applications are in the pilot stage, but with encouraging results—e.g., 15% improvement in tire life, Predicting SAG mill gear failure 3 – 6 months out. The big challenge is to ensure that you can effectively translate the real world into the math world and back to the real world. In our experience, you need excellent operational expertise as well as a data science.

In practice, the problem is defined in Operations World, then solved in Math World, followed by multiple iterations between the two. Data science has many treasures, but it’s critical to manage the integration of the black box into the day to day job of the supervisors.

Want to learn more about extracting value from new technology in mining? Read the full blog: