Technology Showcase 2019: Rio Tinto’s deep-learning AI building on AutoHaul’s success

Perth’s Resources Technology Showcase 2019

Written by Sarah-Jane Tasker, The West Australian

Thursday, 28 November 2019 12:25PM

Rio Tinto’s boss of ports and rail Ivan Vella says the increasing bank of data the industry is generating is the greatest untapped “enabler and disrupter” available to the sector as he reveals expanding artifical intelligence across the global miner’s business.


Mr Vella, Managing Director of Port, Rail and Core Services at Rio Tinto, told the Resources Technology Showcase today that as the miner had moved to remote operations and asset automation it had generated a huge amount of data across its business.


“Today, we track everything, our team is swimming in an ocean of data, which will be crucial to ensuring the efficiency and ongoing health of autonomous assets and systems,” he said.


“Without a doubt, it is the greatest untapped enabler and disrupter available to our industry.”

Mr Vella highlighted in his speech, ‘Project Tempo’, which Rio developed

with EY data and analytics, Monash University and Strukton Rail.


It is an AI solution that can prioritise track condition issues and develop maintenance recommendations that he said returned the highest business value.


"We can now pinpoint problems to within 100 metres on any part of our rail network and the deep-learning AI models predict speed restrictions up to 16 weeks prior to them actually occurring based on track degradation information,” he said.

“While harvesting and analysing data like this is already delivering great efficiencies and improved productivity, we know that the next phase of automation is to get our assets — our trains, trucks, and drilling systems to talk to each other.


“We are now looking at how we can apply this type of advanced AI in other areas of our organisation.”


Mr Vella also shared learnings from the global giant’s use of new technologies, pointing out that AutoHaul — Rio’s autonomous train project that was first conceived in 2006 — cost almost double than Rio first thought and took longer.


“One thing we also know is that large-scale automation and innovation will take longer and cost more than you think,” he said.


“We had to suspend development (of AutoHaul) during the global financial crisis and a 12-year turnaround for a project this significant was not what we had in mind when it started. But it was, and continues to be, absolutely worth it,” he said.


Today, almost all of Rio’s train journeys are completed autonomously.


Mr Vella said the industry was working in a time when the impact of technology, automation, data science and machine learning were offering considerable opportunities.


“I firmly believe that the application of these technologies is essential for our sustainability,” he said.


Mr Vella added that there was no doubt that automation and technological advancements can be a source of anxiety for employees.


“Many in our community and in our companies don’t know where they fit in this changing landscape and I think we have to put our hand up and say that anxiety is in part due to our missteps in the past,” he said.


“During the boom and subsequent downturn, we grew and contracted so quickly that our people didn’t always feel valued or listened to. We have to be better than that.


“We have to be open and honest with our workforce, explaining that no one has a ‘job for life’. But we also need to support our people as they recognise that things are changing and that they need to continue their training, education and self development.


“We all need to put people at the centre of this automation journey, because it simply won’t work if you don’t.”

Strukton_R_logo_CMYK.png

Availability. Reliability. Safety.