Automated geological exploration mapping using convolutional neural structures and data acquisition methods based on swarms of drones

The objective of the study is to develop and validate an integrated approach to improve the accuracy, speed and cost-efficiency of the geological exploration processes. This objective was achieved using the methods of deep learning, swarm intelligence, geoinformation analysis and mathematical modeling. The data collected using a swarm of drones at pilot exploration sites served as the empirical basis of the study. The results obtained show a significant increase in the classification accuracy of geological features (up to 95%), a reduction of the data processing time (by 30-40%) and a decrease in the exploration costs (up to 25%) when using the proposed approach. The practical value of the work consists in creation of a scalable solution for automation and optimization of the exploration mapping processes, which contributes to increasing the efficiency and environmental sustainability of geological projects.