We had the opportunity to discuss with Hugh Agro, President and CEO, of Revival Gold Inc the importance that integrated 3D multidisciplinary models have had in the discovery of new mineralization at the Beartrack-Arnett gold project in Idaho, USA. We discussed Mira Geoscience’s contribution to understanding the project area using modern exploration techniques such as the generation of an exploration model driven from a detailed structural interpretation and Artificial Intelligence, improving unbiased mineralization prediction with a data-driven targeting workflow approach. These techniques proved to be helpful in generating a new 3D exploration model to support Revival Gold’s move into new areas of interest.
Q: Revival Gold seems to be committed to the data-driven approach; why is that?
First and foremost, because we have so much data. The project has a long history, excellent data and information that we want to utilize to the full extent possible. Replicating data at today’s data acquisition cost is very expensive. Secondly, because there’s an unbiased approach that comes from applying technologies like AI and machine learning in this comprehensive holistic approach. Professionals can have bias, geologists can have bias, and our own team is very close to the project. By working with Mira Geoscience we were able to get outside experts, across multiple disciplines, looking at our project with no constraints, no preconceived notions about what they might find, and it’s been very instructive.
Q: Can you describe a little bit more how GOCAD Mining Suite (GMS)1 was valuable in the understanding your exploration data?
Again, it’s the technology. GMS comes out of the oil industry where I first worked early in my career. The oil industry got way ahead of the mining industry in terms of understanding 3D data and visualizing that information. The technology has now moved to such a point where we don’t have to conceptualize just in our minds what’s happening. Today’s sophisticated tools allow us to share, analyse, discuss and collaborate on geological data more efficiently. Working with Mira Geoscience gave us the expertise to drive those tools.
Caption: Detailed Modelling of the Beartrack Zone. Smoothed Au ppm Assay data assigned to the Panther Creek Fault surface.
The detailed model at Beartrack seem to identify 2 major grade trends:
1. Shallow S dipping trend associated to the intersection of the Panther Creek shear and “old” NNE Se dipping structures (yellow)
2. Steeply N dipping trend associated to the intersection of the Panther Creek shear and thrust related structures (green)
© Revival Gold Inc. All rights reserved
Q: Can you describe how AI2 has been used to help outline the area’s potential to host major mineralization?
For us it’s an unbiased approach to identify associations between an area of known mineralization and its geological features, and then applying that knowledge across our entire data set. Our minds can’t think as quickly as computers, nor can they process and manage large volumes of the data. AI takes bias out of the mix. What was very important to Steve Priesmeyer, our VP Exploration, and to our exploration team was to make sure that the process was also knowledge-driven. Mira Geoscience created exploration vectors on which our team provided input to drive learning led by observation. This approach helped us to put everything into perspective and to come up with credible new exploration targets.
Q: Can you describe the benefits of using AI?
I think it’s important. Revival Gold has a very sophisticated board of directors. We’ve got experienced technical advisors who have worked all over the world and a management team that’s worked in multiple jurisdictions on multiple projects and with multiple consultants. It wasn’t by accident that we landed with Mira Geoscience. AI, in the absence of the right driver, isn’t necessarily going to lead us to the right conclusions or deliver for us the right value. That is why we specifically went to Mira, because Jean-Philippe and his team were able to show us that they could bring together multiple geoscience expertise and work seamlessly with our team to pick up our insights on the data. Not just some component of the data; all of our data was put into the one system. It’s one thing to talk about artificial intelligence, but applied AI is another thing. We can no longer only go with the gut feel of the geologist; we must fully leverage the data. Our shareholders want us to be efficient with the money we spend, and we need good business cases for gaining their future investment in our project.
What I like is that Mira Geoscience has the tools and technical expertise in multiple disciplines. We learned through the process that there’s a good degree of practical experience in Mira’s consulting team, which is important because they have to win over the confidence of seasoned explorationist in companies like ours. This method for exploration is a shift, and it’s going to become ubiquitous. It’s come just at the right time because the data in itself costs a ton of money to acquire now and finding new deposits is not as easy as going to find the next green stain on a mountaintop. We’ve got to look more carefully and more cost consciously for new deposits to feed the growing demand for metals.
Q: How has the new detail structural interpretation helped you target high grade mineralization on known deposits?
The actual location of high grade is important, but so are the frequency and the trend of high grade mineralization. For example, the main mineralized trend at Beartrack-Arnett sticks out like a sore thumb but what’s harder to understand is where we get concentrations of higher grade along the structure. Through the work that Mira Geoscience did we were able to visualize intersecting structures and understand how they come into the main mineralized trend to best target higher grades. When you have one trend that is so obviously mineralized, it can distract you from the subtleties and the nuances that lead to the identification of better, or additional, mineralization and better success from your exploration. I think that what we learned from the structural model that Mira Geoscience developed for us is that there are subtle trends in the main system. Mira’s work was also very helpful in giving us an understand what’s happening regionally.
Caption: Targets and mineralized system. Targets are generally associated with high magnetic residual values that could correspond to high heat intrusive centres. The BT mineralization could represent a telescoped system from the East anomaly in a down thrown fault block. © Revival Gold Inc. All rights reserved
Q: Mira Geoscience worked for you in our traditional role (integrated interpretation, investigative modelling and construction of a Common Earth Model). How has this worked for you?
It’s a catalyst for our own group’s thinking. Working with Mira gave us food for thought and facilitated constructive discussion. It helped us to question and refocus our own attention and challenge our assumptions. It’s always good in a discipline like geology (which is part art and part precision) to have debates around the merits of one aspect of the geology versus another. Data acquisition is expensive. Drill data collection in the United States costs upwards of $400 a meter. If we spend on 3D modeling and AI consulting to be able to target our drilling better, it’s far more constructive for our exploration efforts in the long run than to just go in, you know, on a gut feel. It puts us miles ahead.
Q: Mira Geoscience has been helping Revival Gold making a new exploration model. What differences will the new model make to the geological understanding of the area? Have there been any major surprises so far?
There have been a number of surprises that have come to light as a result of Mira’s work. I talked about the structural interpretation and the implications of that in terms of understanding the main Beartrack trend. We’ve also been able to move with a broader scope, to think about clusters of mineralization at Arnett. Instead of [thinking of them] as discrete elements, we’re now seeing them as a cluster of zones, all within a four-kilometre diameter area centred around the Haidee area at Arnett. Again, previously we were looking at each piece just discretely, and as a result of the work that Mira Geoscience has done, we are now able to see the broader trend that centres around the known oxide deposit at Haidee.
With respect to the geological understanding, Mira’s involvement has helped to challenge and improve on our assumptions. Mira’s work augments our own thinking and provides us with a tool that we can use to communicate the geological picture. It’s one thing to have all of this geology in the minds of individuals at the site, but we have to be able to share the knowledge so that others can see the potential. That’s very important to me as the President and CEO of the company, to be able to communicate the amazing potential we have at Beartrack-Arnett. Not just within our geology team, but with outside experts, our shareholders, and others. It is a very exciting deposit. The deposit is now three million ounces and we think as a result of the work that we’ve been doing with Mira Geoscience that there is clear potential for this system to host five million ounces or perhaps more.
1: GOCAD Mining Suite (GMS) is the only 3D interpretation environment that allows the creation of integrated geologic models that honour all geoscientific data, whether the data are sparse or dense. GMS, along with VPmg, enables you to interactively interpret and model geologic data using borehole physical property measurements with potential field survey data. This unique environment, in concert with four different geophysical inversion styles, gives you full control to test geological hypotheses through forward modelling, extend geologic models, or discover new prospective areas.
2: At Mira Geoscience, we are dedicated to developing efficient customized solutions to geological problems using multiple modelling and machine learning approaches. We understand that great emphasis must be put on geoscientific understanding of the data sets prior to the application of machine learning techniques. We offer the right combination of software, geoscientists, and data scientists to solve targeted problems and move exploration forward. The exploration targeting in this case was conducted using a combination of knowledge-driven modelling and Random Forest classifiers. The feature engineering was conducted in close collaboration with Revival Gold. The Random Forest algorithm offers great resilience to noisy data and the resulting predictive model is highly interpretable, providing insight into the key exploration features for a given deposit type.
Hugh Agro, P.Eng. President & CEO of Revival Gold Inc
Hugh is the President & CEO of Revival Gold Inc., a growth focused gold exploration and development company. Prior to Revival Gold, he co-founded Carbon Arc Capital Investments Inc., a private-equity backed investor in mining and metals, and served as Executive Vice President, Strategic Development with Kinross Gold Corporation. At Kinross, he was a member of the Executive Leadership Team and responsible for strategic and operational leadership of Kinross’ growth initiatives including corporate development, global exploration and commercial activities in Russia. Previously, Hugh held senior executive positions with Placer Dome, Senator Capital Partners and in investment banking with Deutsche Bank’s Global Metals and Mining Group. Mr. Agro has served on the Board and Audit Committees of Victoria Gold Corp., Chantrell Ventures and Americas Silver Corp. Additionally, he has served as a Director, Chairman and Interim CEO of Strata Minerals and currently serves as a Director of Fort Berens Estate Winery Ltd., an award-winning winery located in British Columbia, Canada. Hugh holds a Bachelor of Science in Mining Engineering from Queen’s University (1989) and MBA Finance from UBC & London Business School (1997).
Jean-Philippe Paiement, P.Geo.
Jean-Philippe is Mira Geoscience’s Director of Global Consulting. He brings 10 years of mineral exploration experience including expertise in geostatistics, structural, geological, and geochemical modelling and interpretation. He is skilled in the application of machine learning to overcome geological challenges and with new methods to reduce interpretational risks with geological data. He has a wide range of experience in mineral resource estimation for precious metals, base metals and industrial minerals across diverse geological environments around the world. He obtained an MSc from Laval University. Jean-Philippe is based in Quebec-City.