Unearthed Sydney 201620 Aug - 22 Aug
From Idea to Prototype on Resource Challenges in just a Weekend
This competition has finished.
Surges from the SAG mill lead to decreased throughput of the processing operation which cost millions of dollars in lost production. An early warning system which could alert the operator 5 minutes ahead of a surge event would allow them to take control of the system to minimise the impact of the event. SPEWS uses a random forest machine learning algorithm to identify patterns which lead up to surge events. The system can then operate on real time data to provide heads up warnings to plant operators and comprehensive reporting on surge metrics for events which cannot be predicted.