88���ֳ�

The technology, which could significantly reduce machine downtime, was developed as part of an industry-funded PhD project. It is based on a combination of engineering knowledge and neural network technology

Dr Daniel Bongers, a mechanical engineer from the Cooperative Research Centre Mining (CRCMining) in Brisbane, developed the fault detection system, which can determine the state of equipment using existing machine sensors already in place to monitor things such as temperature, voltage and position.

“We can accurately detect faults up to five minutes before they cause the mine to shutdown,” Bongers says. “This forewarning will allow miners to temporarily change what they are doing to avoid any downtime at all.

“The accurate and timely detection of machine faults will significantly increase the profitability of Australia’s largest export industry,” Bongers said. “Continuous production can be conservatively valued at $150,000 per hour.”

Bongers said mines could employ his software as a predictive tool to avoid machine damage and to diagnose faults once they had occurred.

Full commercialisation is anticipated within two years. Bongers and his team are three months into an ACARP project to get the software installed at a mine site. He said he expects it will be implemented at two sites within nine months.

Fresh Innovators is a national initiative to bring the work of 16 early career innovators to public attention.