Skip links

Machine Learning

What is Machine Learning?

Machine Learning combines statistics and computer science to enable computers to learn how to do a given task without being programmed. Just as your brain uses experience to improve at tasks, so can computers.

Let’s have a look at something used widely throughout the world. Google Search. Every time you use google search, you’re using a system with many machine learning systems at its core. From understanding the text of your query to adjusting results based on your interests, to which search results are shown to you first.

Machine Learning and Artificial intelligence is a term that has been advertised across many industries, often overpromising benefits and leaving audiences hesitant. At Cascadia Scientific, our mission is to make Machine Learning understandable and available to operating open-cut mines.

How is Machine Learning Applied in Mining?

Mining operations generate massive amounts of data. As the volume of data increases and surpasses the ability of humans to make sense of it, we will increasingly turn to automated systems that can learn from the data and, importantly, derive meaning. At Cascadia Scientific, we use Machine Learning techniques to derive meaning and insight for mobile mining equipment.

If we use haul trucks as an example, a truck’s operational efficiency is influenced by many factors. These include equipment-specific characteristics such as the age of an engine or the degree of wear on its components; global variables such as the weather and the road network condition; and haulage characteristics that vary between haul cycles, including the truck driver and its assigned workload. So how do we derive value from this data set? We train machine learning models.

A model is only as good as the data you feed it.

Cascadia Scientific platform includes several IIoT sensors to ensure the delivery of accurate and high-quality data. This includes fuel flow meters, accelerometers, altitude and positioning sensing, and vehicle network data. The resulting data set enables Cascadia to train and interrogate powerful statistical models. For haul trucks, we typically employ two forms of models for mining haul cycle analytics: linear regression and gradient-boosted trees (GBTs).

These models deliver insights and actions that can improve nearly every aspect of a load-haul operation. Whether it be Maintenance, Operations, Mine-Design, Forecasting, or Emissions Management – there are plenty of opportunities. View the case studies below to see how mines have benefitted from Machine Learning.

Machine Learning
in Mining

Gradient Boosted Tree Model Training Explained

Examples of Machine Learning Applications in Mining