Earth AI vertically integrates the search for important minerals

A model is only as good as its data, and for Roman Teslyuk, the data wasn’t coming fast enough.

“I hate delays,” Tslik, founder and CEO of Earth AI, told TechCrunch.

Over the past few years, Earth AI has searched for important metals such as copper, platinum and palladium in parts of Australia where no one expected any to be found. The startup’s AI models suggested some locations that proved promising, but locating the rocks with the highest concentration of minerals was slower than Teslyuk would like.

He said the problem was in the laboratories.

“Since we ramped up drilling capacity, we started getting these massive delays,” he said. Laboratories that process rock samples for evidence of important minerals typically have delays of about two months, Teslyuk said. But recently, with increasing interest in developing new sources, delays have multiplied. “We are 7 kilometers behind – 7,000 meters of samples for which we have no data.”

So Earth AI is creating its own labs instead, the startup exclusively told TechCrunch, hoping to reduce the time from five months to five days.

Earth AI models were good at highlighting areas likely to develop into a mine, but once identified, the startup still needed to drill to ascertain what minerals were underneath and how they were distributed, Teslyuk said. Underground mining has come a long way, but there is still no alternative to drilling.

Once the drill cores are extracted, they need to be processed in a laboratory. “We don’t know whether we hit the gold or not. We can’t see it with our eyes,” he said.

For final decisions about the economic value of the mine, including those that may impact a sale, Earth AI will still use third parties to validate its discovery. But during the exploration process, a rapid in-house laboratory has the ability to significantly reduce costs by ensuring that drilling is sent to the correct locations to obtain the best data for the model.

“If you don’t have the answers in time, you have to wait five months to get the answer, the next question [of where to drill] “It’s not as good as it could be,” Teslyuk said. “To keep digging to a minimum, you have to ask the right questions effectively, to get information at the right time so you can narrow down your search to exactly where to go.”