BHP is applying artificial intelligence in practical, operational ways to improve safety, reliability and performance at its Western Australia Iron Ore (WAIO) operations. One example is Oversize and Foreign Object Detection (OFOD), a computer vision and machine learning solution addressing a long‑standing operational challenge.
Oversized rocks and foreign material have been a persistent challenge in iron ore operations, contributing to around 1,100 hours of crusher downtime each year and increasing reliance on higher‑risk manual intervention.
OFOD applies real‑time computer vision and machine learning during loading and along conveyors before crushers and port systems, to identify problematic material early. When detected, the system supports controlled responses so material can be safely removed before causing disruption, shifting risk management from manual controls to engineered controls embedded in the system.
At WAIO, the solution is targeting a 20 per cent reduction in crusher downtime and a 60 per cent reduction in port downtime, delivering approximately 0.8 million tonnes per annum of additional throughput and around US$46 million in value at scale.
By embedding AI directly into operational workflows, BHP demonstrates how intelligent automation can improve safety, reduce downtime and help frontline teams make faster, more reliable decisions in complex industrial environments.
“From improving reliability to supporting faster, better‑informed decisions, these are just some of the ways AI‑supported tools are helping our teams get more value from our assets.”
– Johan van Jaarsveld, Chief Technical Officer, BHP
Image credits: BHP