Just a little under a decade ago, I worked at the mining giant, Rio Tinto. A short time into my tenure, my team was approached by a long-tenured and very senior member of the organisation with a request for a dedicated procurement person to be assigned to a confidential project. The project turned out to be a collaboration with one of Rio’s biggest heavy equipment suppliers on a novel idea: a driverless haul-truck. It was my first exposure to the idea that a vehicle could be autonomous.
Then, as now, iron ore was a big part of Rio’s profitability. From 2003 onwards, China was laying down one infrastructure project after another, and their appetite for the iron ore to go into the steel that would support those projects was bottomless. Rio was looking at more and more ideas to improve speed and profitability, increase worker productivity, and ensure security of supply. By the time I joined in 2007, they had already established their own tyre retread facility with Michelin in order to safeguard against scarcity of the AU$160k tyres for large trucks. Another truck-related cost concern would have been the drivers. Having long used high wages as the bargaining chip against unionisation and to attract Fly In Fly Out (FIFO) workers to inhospitable (and sometimes downright hostile) areas like NW Australia’s Pilbara region, Rio Tinto’s wage bill must’ve been a big reason to examine how to reduce the dependency on humans. I was never close enough to the project to know the underlying reasons so am merely speculating here. Wages, and the high focus the company put on safety of its personnel, would’ve made for interesting bar graphs when it came to examining the places to take cost out of its processes.
Rio’s project must be one of the older ones of its kind out there now. So when The Economist published an article in this week’s edition, I was interested in the findings because I know they are probably the most reliable out there when it comes to the impact of driverless vehicles upon costs. Here are some reported stats from that article:
- Over a 12-hour period, manned trucks are still economical.
- Over 24 hours, or longer, not so much. No coffee breaks, driver fatigue or changes need to happen. The trucks stop once a day for refuelling. And then go on. And on.
- The workforce at the mine in the article is one third lower due to automation.
- The autonomous vehicles are estimated to have run costs around 15% lower than their manned equivalents at the same site.
- Autonomous drilling allows for about 30% longer running than manned rigs at another site.
- Rio Tinto is hoping to win regulatory approval to run autonomous rail cars next year. Their own tests suggest a 6% improvement in average speed and the elimination of three driver changes in every 40 hour shift.
If I was running a fleet right now, I’d be looking to these numbers with a lot of interest. Rio‘s autonomous programme is nearly a decade old and those haul-trucks have been hauling for enough time for these stats to be worth mining.