Caldera Analytics utilizes machine learning to unlock the value of historical mineral exploration data and help discover the next economically viable deposit of new economy minerals. They aim to be the leading provider of pre-drilling machine learning-based solutions for the mineral exploration industry in Australia, by developing models that can be applied to all stages of the exploration process. Caldera Analytics’ deep learning models can adjust their predictions and gain more certainty as more data is collected, allowing clients to make smarter and faster decisions.

Unlocking the Value of Historical Mineral Exploration Data: How Caldera Analytics Uses Machine Learning to Discover New Deposits of New Economy Minerals

Caldera Analytics is a start-up with a mission to revolutionize the mineral exploration industry in Australia by leveraging the latest advances in machine learning to unlock the value of historical exploration data. The company's goal is to discover the next economically viable deposit of new economy minerals, critical for the green revolution, in under-explored areas that are difficult and expensive to explore.

Caldera Analytics' technology addresses the difficulty in exploring for new economy minerals in the frontier areas of Australia. A lot of the 'easy' deposits in Australia have already been found, and the remaining frontiers of exploration occur in under-explored areas that are often obscured by hundreds of meters of irrelevant rocks.

Caldera Analytics' solution is to add a new layer of decision-making information by learning to predict mineralization and other relevant geological characteristics derived from historical exploration data and applying it to these new exploration frontiers. By doing so, companies can make more informed decisions about where to focus their exploration efforts and increase their chances of success in finding new deposits of these valuable resources.

Caldera Analytics has developed a framework for using powerful deep learning computer vision models to make predictions in the mineral exploration domain. These models have previously had implementation issues in this domain due to the lack of sufficient historical training data. By implementing the latest in computer vision research, Caldera Analytics has overcome this constraint by training generalized mineral exploration deep learning models that can be 'fine-tuned' to answer any questions asked by a mineral exploration client, including predicting areas of economic mineral deposits.

These models are smart enough to be used at all stages of the pre-drilling exploration process. If a client collects new geophysical or drilling data, the model predictions can quickly be refreshed to incorporate this new information and further de-risk the exploration project. These new predictions can be generated within minutes, increasing the turnaround time compared to waiting for human consultants to update their interpretations.

Caldera Analytics' vision is to become the leading provider of pre-drilling machine learning-based solutions for the mineral exploration industry in Australia, by developing models that are applicable to all stages of the exploration process. By helping clients make better decisions based on their machine learning products, Caldera Analytics aims to establish long-term partnerships with its clients and become directly embedded in the exploration process.