Rio Tinto, AutoHaul, Cape lambert, iron ore, (L-R) Graham Baxter, supervisor, Lianne Bailey, loco yard driver, Nicole Graves, train examiner, Justin Angove, superintendant, Western Australia.
Rio Tinto workers and one of its robot trains. Three-quarters of the company's trains run without drivers as it aims to slash costs, streamline workforces and increase productivity

When a runaway train carrying 238 wagons of iron ore derailed in Australia this month due to human error and equipment failure, it cost BHP Billiton millions of dollars in lost production.

But it also turned the spotlight on how replacing people with autonomous technology can boost safety and efficiency in the mining industry. 

Rio Tinto is rolling out the world’s first heavy freight driverless rail network in the Pilbara, a remote desert region in Western Australia that supplies half the world’s seaborne iron ore trade.

Three-quarters of its trains run without drivers as they carry ore from its network of 16 mines to ports, from where it is shipped to Asia. The Anglo-Australian miner expects to get close to 100 per cent by the end of the year. 

“We are convinced this is safer than driver-led operations,” said Ivan Vella, Rio’s managing director of iron ore rail, port and core services. 

“Whether it is one, two or more drivers, the controls, safety and the extra things we have done in AutoHaul [Rio’s driverless train system] is safer.” 

Rio is leading a high-tech revolution sweeping the mining industry, which is deploying driverless trains, trucks, drills and loaders that are typically controlled from remote operations centres in cities far from mine sites.

The rollout of autonomous systems has accelerated in the wake of the 2012-2015 commodities crash, which forced global miners to slash costs, streamline their workforces and increase productivity. 

The biggest iron ore miners in the Pilbara — BHP, Rio and Fortescue — say introducing autonomous trucks is generating productivity gains of between 15 and 30 per cent.

A graphic with no description

Automated vehicles controlled by GPS technology are more precise, use less fuel and incur less wear and tear than human driven vehicles.

And the trend looks set to continue, according to consultants McKinsey, which predicts robotics and big data will produce $290 to $390bn in annual productivity savings for oil, natural gas, thermal coal and copper producers by 2035. 

There is a cost, which weighs heavily in the remote mining towns of Western Australia: removing humans saps their life blood.

Some experts believe the industry faces the same kind of dramatic upheaval that the manufacturing industry experienced in the 1970s and 1980s, with robots replacing people in many manual roles. 

“Up to half the 120,000 mining jobs in Western Australia will be impacted by automation and nobody really knows yet if they will be doing different jobs in the same company, area or in a different industry,” said Paul Everingham, chief executive of Western Australia’s Chamber of Minerals and Energy. 

He adds that more blue-collar jobs will be lost than white collar, a view backed up by research from consultancy group BDO.

BHP train crash from videos
The aftermath of the train derailment in Australia earlier this month that cost BHP millions of dollars in lost production

It forecasts half of mining jobs would be replaced by robots by 2020. Half of those miners would be retrained to run the technology in higher-skilled roles controlling the robots, according to BDO’s Near Future of Mining report. 

It added that mining accidents would be cut by 75 per cent by the same year as a result of the switch to robots.

However, there are risks for groups such as Rio as the deployment of robots comes with upfront capital costs and the potential problems of being an early mover, where others learn from the mistakes of the pioneer, in the drive to roll out complex integrated systems.

On the plus side, the moves are providing miners with a treasure trove of data from the myriad of sensors deployed to enable autonomous vehicles to operate. Many are using this information to optimise their mining operations. 

“With AutoHaul we have a lot more data and information about exactly where our trains are, what speed they are going and what path they are taking to which mine,” said Mr Vella. 

A graphic with no description

He points to a bus-sized computer screen at Rio’s Perth-based operations centre, almost 1,500km from Rio’s mines, which tracks the movements of its network of 200 driverless trains.

Separate teams control the movements of trains, trucks, drills and other autonomous equipment, but by working together they aim to provide a seamless supply chain capable of shipping 340m tonnes of iron ore per year. 

The productivity improvements delivered by driverless trains, which Rio dubs the world’s biggest robots, are also enabling it to eliminate driver shift changes, run trains at optimal speeds and reduce wear and tear on rolling stock and track.

This is allows Rio to move an extra 10m tonnes per year and interrogate the data to change how it runs the rail network to drive further efficiency gains, says Mr Vella. 

Safety enhancements introduced with AutoHaul include new collision detection and obstruction detection systems, as well as a suite of controls that kick in when onboard systems fail.

A graphic with no description

Rio says drivers no longer have to make 1.5m kilometres of car journeys every year to board trains across its 1,700km rail network, reducing the risk of accidents. 

On top of this, Rio is planning a $2.2bn investment in Koodaideri, its first intelligent mine that integrates autonomous equipment, machine learning and cloud-based internet technology in a move that will place more decision-making in the hands of computers. 

“This puts all the pieces of the puzzle in one place,” said Mr Vella. 

“We will create a complete digital model of the mine, which allows us to mimic the real world physical asset and interrogate, model and optimise it. All the sensors at the site are wired and so we can get machines to talk to machines to improve the way they interface with each other.” 

Unions warn of damage to communities

Introducing driverless trains has not been without challenges.

Rio Tinto’s project is running four years behind schedule and has cost $940m — almost double the budget announced in 2012.

The miner blames software issues and additional testing for the delays, which forced it to reduce its iron ore production forecasts last year. 

AutoHaul is also causing concerns among Rio’s 420 train drivers, among the highest paid in the world with earnings up to A$250,000 a year, as well as officials in small mining towns in the Pilbara that rely on mining jobs for their survival.

Frustration caused by the uncertainty created by the driverless train project was one of a number of factors that prompted trade union members to vote down an enterprise bargaining agreement in July. 

“Drivers are uneasy because they don’t know what the future holds for them,” said Greg Busson, secretary of the CFMEU union in Western Australia. 

He said the union is not opposed to AutoHaul but wants Rio to take more account of how it is impacting employees and communities in the Pilbara region. 

Unions warn communities may not be as willing to accept mining operations in their neighbourhood if most jobs are relocated to cities and there are less local benefits.

Highly automated and remote-controlled mines will have a harder time getting approval by local authorities and are more likely to face restrictions in their operation, they warn.

Rio says there will only be a modest reduction in train drivers due to AutoHaul, possibly 10 per cent, as it still needs people to supervise the trains and drive on some portions of its rail network, such as yards. The company says it is reskilling and redeploying workers displaced by technology.

Jamie Smyth 

Get alerts on Driverless vehicles when a new story is published

Copyright The Financial Times Limited 2021. All rights reserved.
Reuse this content (opens in new window)

Follow the topics in this article