The thing about mineral discoveries, as we say in the industry, is that it’s generally more a matter of luck rather than scientific prediction. The modern individual explorer has a lot of techniques available to them. While many may not be new, computers have allowed us to improve precision, and provide us with the ability to overlay large and different datasets in order to look deeper for the ever-elusive hidden ore. You could say that discovery is a result of diligent application of every tool available to us, but especially datasets.
It is known that tech giants like Google, Facebook and Amazon reap the fruits of using big data to learn about users’ behavioral patterns. Experts in our field have started experimenting with big data in order to push the boundaries and discover new and profitable areas of mineral deposits. In an effort to learn more about this field, these experts have started applying advanced data science techniques to mineralogy. The Deep Carbon Observatory published a paper in 2017 that discusses how ‘network theory’, usually used to analyse things like a spreading disease, or terrorist networks, can actually be used to identify mineral diversity and distribution across the globe.
What is big data?
Let’s break this down: big data is something that large organisations like Netflix and Amazon use to do a market analysis. The idea behind it is to be able to predict consumer habits; and now mining companies are jumping onboard to do the same. It’s basically taking a mathematical theory that was previously used to identify spending habits of customers, to look for mineral deposits, and even the formation and existence of minerals that science may not have even discovered yet.
Dr. Shaunna Morrison, from the Carnegie Institute of Science, addresses that this practise of using big data is now spreading across the mining industry: “[they] are being implemented across the board by all these huge companies. They’re very powerful statistics.” And, she says, just like the spending habits of consumers re-occur across websites like Amazon, so does the formation of mineral across the Earth, so it makes perfect sense to characterise them similarly.
Data = Power
The way these companies have been using this big data is pretty interesting. Firstly, they’ve formed databases where they’ve put in all the relevant information about evolution data, geographical characteristics, how often mineral re-occurrences have been found in specific areas and their trends, age of the minerals, and lots of other information. From there, a group of experts can then analyse the data to look at similar locations that will have the highest probability of containing a large mineral deposit. This technique also lets the scientists in the industry visualise new and unique patterns of distribution and occurrence of minerals. It’s basically allowing them to form a ‘checklist’, that will help predict where minerals are, where they aren’t, and where to go next.
It’s important to pop this information into a database, because we’re not just dealing with the characteristics of one or two geospatial locations. We could be looking at well over a thousand, which is overwhelming for a human brain to handle and analyse by looking at a spreadsheet. In fact, even the algorithms that are being used can gauge where these deposits might be, but they can’t predict anything about size or quantity. The answer to this problem? Obtaining more data.
Dr. Morrison states that the issue is currently the lack of data available on “the amount of materials on Earth’s surface right now”. An issue leading to that is that the mining industry has a lot of information regarding this, but they’re not willing to part with the knowledge. Dr. Morrison believes that a part of this is the limited understanding that mining companies have of the data they hold. Mining companies dedicate one percent of their IT to big data, compared to five to seven percent for most industries.
Is data more important than people?
All this data doesn’t have much use without the people behind it creatively interpreting and testing it. Case in point; the Goldcorp Challenge in the year 2000. For a total prize money of CAD$105,000, anyone worldwide could form teams to help Goldcorp find more gold at their Red Lake mine. They based this information off of a wealth of proprietary data that Goldcorp was publishing on public domain for free for the sake of the competition. If nothing came out of it, Goldcorp had just released a veritable treasure trove of data, in the hopes of finding something against the odds. But luckily, that wasn’t the case.
There was a winning formula in the challenge – the fact that the competition was open to anyone, meant that the people who applied were incredibly diverse in their education, work and cultural backgrounds. While there were geologists scattered throughout the competition, it was a blend of people that allowed creativity in interpreting the data provided. Of the top five entries, four were drilled. All four struck gold. And that’s not all – of the 110 sites that were identified in the course of the competition, 50 per cent of them were unknown to the company, and 80 per cent yielded significant gold reserves, exceeding $6 billion in value. Rob McEwen, Founder and former Chairman and CEO of Goldcorp at the time, estimated that just by holding the competition, they’d managed to shave of up to three years of the company’s exploration time.