To see what the future of mining looks like, you don’t have to drop down a cramped shaft into the dark belly of the Earth. You don’t have to crouch inside a cramped mantrip, or follow a conveyor belt along a cold longwall of coal. No, to see the future of the industry, you only have to step into a comfortable, carpeted office in Brisbane, Australia, and look at the massive interactive screen that fills its wall.
Here, in Rio Tinto’s state-of-the-art Processing Excellence Centre (PEC), you’ll see how big data is transforming the company – and the global mining industry – by driving efficiencies, reducing costs and processing tons and tons of raw data.
The geeks in the Rio Tinto offices – at the moment, the PEC is staffed by just 12 mineral experts – are calling this room the ‘Excellent Centre for Excellent Excellence’, a claim that’s really hard to argue with. The PEC’s giant monitor displays complex technical data in real-time, receiving data just 100 milliseconds after it’s produced at the site hundreds (even thousands) of kilometres away.
That data – which comes in at a rate of about 30 GB a day, or a terabyte a month – is picked apart and processed by 20 different analytical systems, providing processing solutions and initiatives for Rio Tinto mine sites in Mongolia, the US and at five locations across Australia.
This is number-crunching like we’ve always done it. But now it’s on an unprecedented scale, and at an unprecedented pace. While unveiling the room to the Australian media, Rio Tinto’s head of innovation, John McGagh, remarked that: ‘Once we had data streams. Now we have data waterfalls. Tomorrow we will have data oceans.’
What makes the PEC so, well, excellent is its real-world, real-time, really effective use of big data. That’s the big buzzword across many industries right now, and – as tends to happen with buzzwords – every expert has their own definition for it. Paul Morgan, technical director at Sandton-based Decision Inc, sums it up as well as anybody could.
‘Big data is an all-encompassing term that refers to very large data sets that cannot be processed in acceptable time frames using traditional database technology. There is no industry standard as to what encompasses “very large” data sets, and the measure of what constitutes “big” is largely subjective,’ he says. And big doesn’t necessarily mean better – as several IT experts are quick to point out.
Big data. That’s the buzzword across many industries right now, and – as tends to happen with buzzwords – every expert has their own definition for it
André Stürmer, CEO of commercial credit bureau Inoxico, highlights one of the sobering drawbacks that come with all those oceans of data. ‘Big data in itself offers little value,’ he says.
‘Without beneficiation through prescriptive and predictive analytics, it is little more than a repository of raw data. For the most part, it is predictive analytics that are of most interest to decision makers as they provide insight and a “precognition” of future events that ordinarily would have been pure estimation.’
For the mining industry – with its age-old heritage of digging through dirt to find the golden nuggets within – this should be second nature. That’s exactly where Rio Tinto is getting it right. Instead of simply gathering vast volumes of raw data, the PEC is taking that data and turning it into something useful.
Much of the early buzz around big data has been around its uses as a marketing tool – assembling massive databases of information about clients and potential clients. However, that assumption is quickly changing, says Raj Wanniappa, Dimension Data’s big data general manager. ‘Through understanding customer behaviour better and through targeted marketing, the opportunity to streamline operations and decrease costs also exists.
‘This is most relevant for the manufacturing [FMCG], resources [mining, oil and gas], agriculture and government [healthcare, municipalities] sectors.’
‘The potential trap is to harvest data that has little value,’ says Peter Searll, managing partner at Dashboard Marketing Intelligence. ‘Focus [on the problem, issue or data needs] and selection of the most appropriate way to collect the data, taking channel independence and privacy into account, are key. Often clients are sold data collection solutions that are sexy, but are not designed to do the job at hand.
‘The insight value lies in collecting accurate, cost-effective and timeous data, regardless of which collection tools are used.’
Morgan says that data in vast quantities has been around for many years. ‘But the technology to handle this volume adequately – especially in real-time – has been expensive and difficult to implement. The critical need is to match business requirements with architecture.
‘Too often it has become a case of using technology for its own sake instead of reflecting on what the goals of the organisation are and what needs to be done to achieve them.’
Across Africa – and in South Africa, where the mining industry seems to roll from one crisis to another – many mines are being forced to cut their IT spend as they try desperately to reduce costs. This, in turn, is forcing mining operations to take a far more strategic approach to technology.
As AccTech sales director Marc Gower told ITWeb late last year: ‘The oil and gas sector uses technology with finesse, where the mining industry has always used it with brute force.’
Mind you, ‘brute force’ comes with the territory in the mining industry, where the on-site equipment and machinery is often (and literally) run into the ground.
Global computer giants IBM believe they have a solution to that universal problem, and again, it involves using big data. Speaking recently to VentureBeat, IBM Research’s manager of smarter planet modelling and analytics Matt Denesuk explained how big data could transform the ZAR50 trillion business of operating mining equipment.
‘The oil and gas sector uses technology with finesse, where the mining industry has always used it with brute force’
Denesuk says that, while IT spending in the natural resource industry is relatively low, safety concerns have already seen most heavy mining equipment being fitted with sensors that broadcast data to communications hubs, where the data is stored in monthly reports and later analysed to see how it’s holding up.
Now, using big data services, IBM is making those analytics available in real-time, allowing mine operators to make detailed predictions about specific pieces of equipment.
Consider the cost of, say, having an excavator go down in the field (IBM offers an estimate of US$5 million a day), or of losing a haul truck (US$1.80 million a day). As Denesuk points out, preventing that equipment from failing in the field would have a massive impact on the mine’s bottom line. He says: ‘A component might be chronologically young, but its excessive usage might have given it a hard life.’
‘There are so many different components tracked and so many modes of failure,’ says Denesuk. ‘The most important are the size of the loads, wear and tear, and the amount of fuel, taken together. But it’s not one thing. It is a combination of factors.’ That’s where the big data aspect comes into it. For its predictive monitoring of mining equipment, IBM has to run about 200 000 streams of data through its complex algorithm.
‘People have been talking about big data for 10 years,’ Rio Tinto chief executive Sam Walsh said at the media launch of the PEC. ‘This is physically seeing it on the ground and leveraging it to look for anomalies and to look for variations in processing.
‘We are seeing the benefits already. In the [past year] we have picked up AUS$80 million in improved cash flow through this centre, so that’s a significant payback already.’
Incidentally, that AUS$80 million cash flow boost saw the PEC paying for itself within a year. No wonder that ‘payback’ is making the rest of the mining world sit up and take notice.
