News

Canadian miner Integra Gold names finalists in crowdsourcing challenge

TORONTO – Could crowdsourcing, artificial intelligence and virtual reality play a prominent role in the evolution of the mining industry?

The 32-year-old chief executive of Quebec-based Integra Gold Corp., Steve de Jong, thinks so.

After acquiring the Sigma-Lamaque mine in Quebec in 2014, the junior exploration company opted to do something considered unusual in the mining industry.

The company took the six terabytes of data spanning more than 75 years of exploration and development at the site and threw it up on the Internet for all to see.

De Jong — by offering $1-million in prizes — was hoping to crowdsource an answer to the following question: Where is the next large deposit of gold at the mine?

The challenge drew 100 submissions from 1,342 individuals hailing from 83 countries looking for a piece of the prize. Artificial intelligence, virtual reality and a geographical computer system known as GIS were among some of the techniques used in the submissions.

Integra named five finalists on Wednesday, among them a Quebec-based team of six post-graduate university students calling themselves the Data Miners.

The team’s submission used machine learning algorithms — a component of artificial intelligence — to identify potential gold deposits.

Vincent Dube-Bourgeois, a member of the team and a masters student in geology, says using data to locate a deposit requires processing more variables than the human mind can handle.

“We needed something that was easy to use,” Dube-Bourgeois said.

“Data is very important in the mining industry, but it’s never published like that,” he says. “Normally everything is so private because there’s so much money involved. Nobody wants to get that information stolen because you pay so much for it.”

Another of the finalists, Quebec-based consulting firm SGS Geostat, also employed artificial intelligence to identify the probability of striking pay dirt in various areas of the site. The team then used an Oculus Rift headset to experience the three-dimensional models it had built in virtual reality.

“Normally we just look at it on a computer screen,” says Jean-Philippe Paiement. The problem with that approach is that it only allows the viewer to experience the model from one vantage point.

“Getting into the model with virtual reality allowed us to get a different perspective, a different point of view. … We were able to really get into the model and look at it from different angles. That helped to validate the drilling targets we produced.”

Paiement says technologies such as artificial intelligence and virtual reality hold great promise for the mining industry.

“Mining companies tend to generate a lot of data and they rarely use of all of it because it’s too hard to correlate 12 different variables together,” he said.

“By using the machine learning techniques we’re able to harness the power of all those variables an find correlations that wouldn’t be possible to find in other ways.”

The other finalists include a team of Australian and Ghanaian geoscientists called Goldcrushers, a group of researchers from the University of British Columbia called GoldRX and Australian data analyst Paul Pearson.

The winners will be announced at an event held in Toronto on March 6 as part of the Prospectors and Developers Association of Canada convention.

De Jong says it will likely take a couple of high-profile success stories before techniques like artificial intelligence and virtual reality become widely embraced by the mining industry.

And, he adds, the technologies will likely be used alongside — and not in place of — traditional and proven geological methods.

“Some people get confused and think that we’re trying to replace one with the other,” says de Jong. “The most exciting thing is to see these new technologies married with proven geology. Because no one’s trying to discount proven geology.”

Follow @alexposadzki on Twitter.