Rio Tinto uses algorithm to reclaim train capacity

Rio Tinto is hoping to unlock an estimated 3 million tonnes of latent transport capacity in its rail network using an algorithm that is currently under test at its Brockman 4 mine in Western Australia.

A Rio Tinto train in the Pilbara (courtesy: Rio Tinto)

The mine giant is developing a system to estimate in real-time how many tonnes of iron ore are loaded into the rail cars that haul ore to port.

General manager of innovation - human factors in Rio Tinto's technology and innovation division, Andrew Shook, told the miner's own magazine the company wanted a better system to prevent the wrong amount of iron ore being loaded into the rail cars.

Currently the rail cars are weighed on the track, but "the scale's results do not arrive in real-time," Shook said.

"There's a delay of four [rail] cars, which can be as much as two minutes, so the [loading] operators don't know if they have put too many tonnes in a [rail] car until it's too late," he said.

"If they have overloaded a car, they have to stop the train and remove the extra mass using a digger, which slows down the whole operation.

"To avoid overloading, operators can err on the side of caution and tend to put less mass in a railcar than they could – in other words they might not push too close to the upper limit."

Rio Tinto is hoping a new system that uses existing sensors and control hardware "but more sophisticated mathematical processing" will ensure rail cars leave mines with an optimum load of ore.

The system estimates in real-time how many tonnes have been loaded into the rail car, produces an "aim load" for the car using statistical analysis, and also contains an "automatic overload controller" to prevent overweight loads.

However, the operators would have final say on how much ore was loaded into each rail car, Shook said.

The system is expected to be live at Rio Tinto's Brockman 4 mine "in the coming months", before being "progressively applied" at other sites.

Already on the rollout roadmap is West Angelas, a test site for driverless trucks and other automated systems under Rio Tinto's 'Mine of the Future' program.

VirtualEYES moves into trial

Rio Tinto also revealed progress on a big data visualisation tool that iTnews first reported was under development in March last year.

VirtualEYES, as it was codenamed then, has been renamed RTVis, but it's role is the same: to "create a rich 3D display of the pit environment that can be easily and quickly understood" by operators.

"The system can be used to identify geological boundaries and provide near-real-time operational context for personnel physically located away from the active mining area," Rio Tinto said.

"It also supports control of automated equipment, such as the automated drilling system in Rio Tinto iron ore."

The system is being trialled in four Rio Tinto product groups: iron ore, copper, energy and diamonds and minerals.