, , , , , ,

Mineral Exploration in Brazil with AI… the story behind the Map

Since the days of the first Gold Rush, when a prospector would head off into the bush with a shovel, a gold pan, and a compass, maps have played an essential role in mineral exploration. Maps allow prospectors, geologists and miners to better understand the world beneath their feet… “why is there a deposit here and not there?”

Today, Minerva Intelligence (TSXV:MVAI), a Vancouver-based artificial intelligence company, uses cognitive AI to build maps for mining and mineral exploration companies that continue to ask –and answer– those questions. Minerva’s maps are developed under the banner of a product called TARGET, an AI-powered prospect generation software tool, to identify locations that have similarities to known mineral deposits. In other words, TARGET finds locations that have the same attributes as those that geologists look for in complex geologic systems.

Earlier this year Minerva Intelligence announced the successful implementation of its TARGET software by Giga Metals Corp. (TSXV: GIGA) in conjunction with the recent release of their Parnaíba Basin Project results. Giga had enlisted Minerva to identify and evaluate new prospective exploration targets in Brazil by building a TARGET map for most of the country.

After validating the results produced by Minerva’s software, Giga Metals made the decision to acquire exploration permits covering significant regional sediment-hosted copper anomalies along the southern perimeter of the Parnaiba Sedimentary Basin in southern Piaui State of Brazil’s Northeast Region.

Minerva was able to build this map by compiling spatial datasets from across Brazil. These geological datasets came from both private and public sources, and represented lithology, mineral occurrences, and geologic terranes, among others. A significant challenge was posed by the need to re-structure the Brazilian data in a manner that would be usable by Minerva’s AI system. This included translating massive country-scale datasets from Portuguese to English, then to GeoSciML-compliant terminologies. Once the data had been harmonized both terminologically and topologically, the data was aggregated using a hexagonal grid approach. Using hexagonal gridding reduces sampling bias from edge effects and better represents the curves of naturally occurring phenomenon, such as lithologic formations.

After processing all the data, over 65,000 hexagons were generated across the area of interest, all of which were compared against 80 mineral deposit models using Minerva’s proprietary reasoning technology. This involved more than 5 million comparisons, and the results were delivered as different map layers for each deposit type. The end product of this analysis was a list of AI-produced target areas throughout Brazil that were completely auditable and explainable, and, most importantly, actionable by Giga Metals.

Using TARGET’s mapping technology, Giga Metals was able to determine that the project in Brazil had a high likelihood of success and, as a result, Giga decided that they should pursue investment in the region.

Giga CEO Mark Jarvis was enthusiastic about the Brazil TARGET project: “TARGET enabled us to wade through mountains of disparate data to quickly and easily find new drilling targets, a result that would not have been possible even a few years ago.”

The results have been released publicly for select deposit models and is available on http://brazil.minervaintelligence.net/target-maps.

Minerva TARGET Web map

“The successful deployment of our TARGET software highlights the value we are able to provide to companies that are managing large datasets and seeking to incorporate an artificial intelligence element into the decision-making process,” said Scott Tillman, CEO of Minerva Intelligence. “Our success with Giga in Brazil, in conjunction with our recent success in Mexico, points to even greater success in the future in delivering results for mining and exploration companies around the world.”

More information on Minerva’s TARGET technology can be found at http://minervaintelligence.com/TARGET.

Related Articles

From flames to data

From Flames to Data: How deep learning can help in wildfire suppression?

We are driven by our survival tendency. And that compels us to know or be able to predict the future.…

International Digest Sept 11

International Geospatial Digest September 11th, 2023

Planning Landmine clearances with UAV LiDAR Mapping pollution from space Tracking dangerous chemicals in the Great Lakes with maps New…

International Geospatial Digest June 19th, 2023

Private Industry, New Tech, challenges ahead for Geospatial Intelligence Early signs of El Niño spotted by International Sea Level Satellites…