Identifying Drill Targets Using Big Data
The CEO of Orefinders says junior companies can use artificial intelligence and machine learning to identify drill targets. With ever greater computer power, Stephen Stewart expects a transformation in the exploration sector.
Stephen Stewart, the CEO of Toronto-based Orefinders Resources Inc., has confirmed that he will be presenting at Mines and Technology Vancouver in June. We caught up with him to hear in advance his views on recent technology innovations and how this might affect the gold-exploration sector.
Can you tell us a little bit about your upcoming presentation? What can delegates expect to learn?
I will be discussing how companies can use data to gain an edge; ultimately how to leverage existing data, knowledge of geologic formations and advanced analytics to improve exploration success. Delegates can expect to learn when, and how, big data will create value in exploration and development of a project, and a discussion on the overall return on investment of using big-data analytics.
What new technology innovation have you seen over the past 12 months that most excites you, and why?
More and more companies are using artificial intelligence and machine learning to generate targets. Big data has always existed, but computing power keeps improving. Goldcorp originally used IBM's Watson to generate drill targets at Red Lake, and the company has seen a solid return on its investment when looking simply at the time saved by their geologists.
What has changed?
Computers can process large quantities of data much faster than humans can, and now we see a few juniors engage in this technology in one way or another — Goldspot Discoveries and Albert Mining are two that come to mind that use machine-learning algorithms or artificial intelligence to generate exploration targets. This technology is exciting, and if proven to work in terms of ultimately mitigating risk and saving costs for junior mining companies, it will be a game changer.
What can the mining sector learn from other industries when it comes to technology implementation?
The mining sector has consistently evolved through innovation. These innovations are not yet on the radar of generalists as typically they are constrained to a mine site and their primary users fall under a specific group of individuals (primarily mining engineers and geologists). Hence, these innovations are not large scale and in the public eye because they do not apply to the general public. That being said, more companies should turn to innovative methods involving big data and artificial intelligence as long as the rationale and return on investment exists.
When it comes to implementing mining innovation, what is the biggest challenge you have faced, and why?
Our priority is quite simple: Return on Investment. We are lucky to be in Canada, and specifically focusing on the Abitibi Greenstone Belt where a lot of the data — over 100 years of exploration and production information — has already been collected and is in one way or another, publicly available. That being said, we have to ask ourselves constantly whether a specific innovation would provide tangible results that will ultimately reduce exploration risk enough that it merits capital allocation in a bear market.
We are lucky to have a knowledgeable, data intensive technical team that works with big data to come up with tangible results. The biggest challenge, however, is to rationalize investment in a still relatively unproven technology when inflow of capital into this industry is still limited. Until this technology is credited for a major discovery, it will not be considered a genuine game changer.
To what extent, do you think, does the price behaviour of gold effect the speed and/or direction of these innovations?
The price of gold is very important. A higher gold price would ultimately reduce the cost of capital for companies levered to the gold price, and ultimately enable them to engage in more exploration. A company would be able to fund such activities without having to dilute shareholders significantly. The real question is whether companies would rather spend the funds raised on drilling or for implementing such technology to generating drill targets. For us, we are generating drill targets using the large quantum of data we have, so that when the market turns, we will be ready to take advantage and drill — which we know is capital intensive. In a bear market, we like to use the data we have and innovate; in a bull market, we like to raise capital to drill test these targets.
What role do exploration companies play generally in the speed and/or direction of these innovations?
Exploration companies are very important in the direction of these innovations but face a lot of obstacles — the biggest being resources and capital. These innovations can't be adopted by exploration companies unless they have sufficient capital to take the risk of adopting an unproven technology, or an in-house team that is capable of handling big data, or are familiar with machine-learning algorithms and artificial intelligence.
As already mentioned, unless this technology is credited for a major discovery, it will not be considered as a game changer — and this discovery does not need to be by a junior company. Going back to the 1950s, it was not until the discovery of the highly profitable Mattagami Lake deposit by Dominion Gulf, which had conducted an aeromagnetic survey, that geophysics was considered legitimate. Now every junior uses geophysics as a major contributor to their overall exploration activities.
Stephen Stewart will be presenting at the upcoming Mines and Technology taking place in Vancouver, 4-6 June 2019.