Rio Tinto IT tool unearths millions in prized iron ore

By on
Rio Tinto IT tool unearths millions in prized iron ore

But "couldn't care less about data".

A 3D visualisation of Rio Tinto's mine sites coupled with sensors on its trucks and processing plants is reaping the miner an extra 900,000 tonnes of iron ore and millions in revenue every year at just one mine site.

RTVis, developed in 2012, is a 3D modelling tool allowing Rio Tinto operations centre staff to visualise in real-time data generated from sensors located on its trucks, drills and other mine equipment.

The tool visualises data located within Rio's mine automation system (MAS), which uses information generated from mine equipment to schedule work and relay these instructions back to the trucks and drills on Rio's mines.

In a speech to the Brisbane Mining Club today, Rio Tinto head of technology and innovation Greg Lilleyman described the significant impact the RTVis tool had made on the miner's bottom line.

Lilleyman said the miner had increased saleable ore product by 2.5 percent and delivered 900,000 extra tonnes of product last year at its West Angelas mine alone.

The price of iron ore currently sits at US$55.52 per tonne.

"You can do your own sums on the value this has delivered," Lilleyman said.

It marks a huge improvement on the $500,000-odd of extra value the company was deriving from RTVis a year after it debuted, thanks to both additional ore and waste minimisation.

RTVis is now used by almost 1000 Rio Tinto employees across 80 percent of its open cut sites around the world, Lilleyman revealed.

"Wherever it goes the gains multiply. Any new features of the software can be rolled out to all sites around the world in a matter of days," he said.

At the Perth Hope Downs 1 mine, Rio Tinto was able to reduce the cost of each drill pattern by around 13 percent or $150,000 within six weeks of implementing the tool, Lilleyman said.

At its Kennecott mine in the US, RTVis reduced long queues and wait times at shift changeovers by allowing incoming operators to pinpoint exactly when their truck will arrive for changeover. This has resulted in an extra 16 loads for the first two hours of each shift, Lilleyman said.

"Visualisation technology allows us to differentiate the wobbly, asymmetrical boundaries between high and low grade ore deposits in three-dimension," he told the conference.

"It brings precision in place of the ‘drill, blast and hope for the best’ approach."

"Couldn't care less about data"

Despite the gains Rio Tinto is making from its big data play, Lilleyman said the miner "couldn't care less" about the raw data itself.

The "sheer volume" of data Rio Tinto is generating - each one of its 900 haul trucks is fitted with more than 200 sensors, and each truck generates 5TB of data each day - is not valuable if not used and visualised appropriately, he said.

"It’s the knowledge we pan from it that is precious. And this information resides in patterns," Lilleyman said.

"The patterns are discoverable, but only with vast computing power and world-class expertise in advanced analytics and diagnostics.

"[And] we transform the information into tools to help our people on the ground make better decisions, all day, every day. It’s those people and their ever-improving decisions informed by powerful new insights that generate the value and continuously sharpen our competitive edge."

He cited the example of Rio's predictive asset health system, which captures, cleans and processes large quantities of equipment data "using a set of rules developed through machine learning and advanced analytics".

It means the miner can quantify the risk of impending equipment failure, and has already allowed the company to extend the life of a haul truck engine from 25,000 to 30,000 hours, Lilleyman said.

Earlier this year Rio Tinto decided to open up its exploration dataset to smaller miners to spread the risk of new projects and identify new opportunities as mineral prices remain under sustained pressure.

Copyright © iTnews.com.au . All rights reserved.
Tags:

Most Read Articles

Log In

Username:
Password:
|  Forgot your password?