Sustainable Cu Extraction

Skouria is tackling an urgent challenge: making copper more affordable. Their new extraction technology promises to cut traditional mining’s environmental footprint, while boosting Cu supplies for green technologies like wind turbines and electric vehicles.

Although Skouria’s experts have studied copper geochemistry for years, selecting the exact right chemical conditions for efficient Cu extraction involves balancing many variables, such as redox potential, temperature, and numerous reagent concentrations. Typically, humans run candidate conditions through a chemical simulator - one-by-one. It’s a slow process where scientists manually enter conditions, and select the next experiment based on intuition.

In May 2025, Alkali partnered with Skouria to develop an AI solution to efficiently identify process conditions for Cu extraction. OptoMine, Alkali’s AI-driven module for mining simulation, efficiently explores the high-dimensional space of Cu extraction conditions, and identifies optimized conditions for Cu extraction.

OptoMine: AI-Powered Experimentation

OptoMine is a platform technology that integrates with various chemical simulation engines. In this case, we integrated with a traditional mining simulation engine. Inputs to the platform include:

1. A base simulation the AI can learn from

2. Experimental bounds to explore

3. A goal to optimize - here, copper extracted

Step 1: Ingest a Base Simulation

Start by dragging and dropping a working simulation file into the platform. OptoMine identifies any adjustable parameters from the working simulation, which acts as a starting point.

Step 2: Select Exploration Bounds

For each identified variable in the simulation file, select a fixed value, or range of values you want the system to explore. You can also fix the ratio of certain variables.

Step 3: Run Experiments

Finally, the experimental campaign is run. AI-driven methods, such as Bayesian Optimization, identify promising experimental conditions which are run through a traditional deterministic chemical simulator. The algorithm learns from its past experiments, to iteratively select more promising extraction conditions.

At the end of a campaign, an optimized set of Cu extraction conditions is surfaced.

For each experiment, a simulation file is saved, so that a human engineer can go back to run, and verify, OptoMine’s results.

Each scatterplot point represents a full chemical simulation. Results are color-coded to show best points, and can be visualized in 2D slices for easy interpretation.

Skouria Results

OptoMine was able to intelligently explore a high-dimensional space, executing 100 experiments in under six minutes–by comparison, it took our human engineer over an hour to execute only 20 experiments. Compared to random sampling, OptoMine identified conditions that yield 5x the Cu concentration, while acting faster than any human.

AI For Process Design

AI doesn’t just have data-driven pattern recognition and fast action–a big advantage in using AI for chemical process design is that it doesn't get tired, which allows engineers to run more experiments.

For example, perhaps we want to explore conditions that optimize Cu extraction for a new type of Cu ore. Rather than a human painstakingly running experiment-by-experiment, one can simply input a new starting ore, and run the algorithm in a matter of minutes.

Contact us if you are interested in bringing OptoMine to optimize your chemical process. If you want to see how Alkali can help with chemical projects and equipment selection, check out our recent blog post on ProcessMate.