Grabuma CEO Veriyadi had the opportunity to speak at the 2025 Mineral Resource Evaluation Conference organized by the Geological Society of London in London, United Kingdom. Veriyadi introduced a novel approach to grade estimation that improves upon both geometric and geostatistical methods. In geostatistical methods, estimation accuracy depends on the spatial correlation between samples and their relationships in estimating values at specific locations. Geostatistics is widely used due to its ability to produce smooth and accurate estimates using Kriging. However, geostatistics performs well only for linear and stationary data. For non-linear and non-stationary data, Kriging can yield inaccurate results. Artificial intelligence approaches, such as artificial neural networks (ANN) combined with metaheuristics, offer a solution to overcome the limitations of Kriging. This method utilizes variables related to nickel content, such as Co, MgO, Fe, and sample location. The ANN learns data patterns that can then be used to estimate grades at target locations. This method is optimized using the Archimedes Optimization Algorithm (AOA) and is known as a hybrid ANN-AOA approach. The method was validated using training and testing datasets. Estimation results using ANN-AOA proved to be more accurate compared to ordinary Kriging (OK), inverse distance squared (IDS), and ANN alone. Therefore, ANN-AOA shows great promise as an accurate content estimation method for complex samples
