The Client

A large mining company wanted to improve their performance and reduce costs.

Mineral processing operations are difficult to control for naturally occurring and variable ore and rock sizes. High-capacity SAG minerals grinding and processing plants reduce ore material to a controlled particle size.

Many different equipment and process variables need to be sensor monitored to produce metrics, in this case every millisecond.

The Challenge

Cause and effect between the monitored variables in mill processing is notoriously difficult to understand and control for i.e. to identify which measured variables are the prime causative ones that most impact others in the overall process so that intervention can occur most effectively at the key spots to improve performance and reduce costs (electricity in this case).

The Solution

Sensor data had been monitored for 4 weeks at millisecond intervals. This was analysed by our Causal AI Engine to identify the cause-and-effect relationships among 25 performance-related variables.

The Results

BDC’s Causal AI Engine detected that one variable had a significant measured causal impact on the performance levels of nearly two thirds of the other variables.

The company was able to understand the particular sensor-driven points in the overall process and took engineering steps that led to increased throughput and reduced costs.