AI-assisted 6G-IoT system for environmental monitoring and risk management in mining sites

This paper presents a 6G-IoT-enabled framework for environmental monitoring and risk management in mining sites, built around a terahertz gas sensor integrating graphene, black phosphorus, and vanadium dioxide. The proposed structure is rigorously analyzed using the Wave Concept Iterative Process, which provides accurate and computationally efficient full-wave modeling of multilayer planar devices incorporating isotropic and anisotropic two-dimensional materials. The formulation is validated against analytical results, confirming both its accuracy and its significantly reduced computational cost with respect to conventional approaches.The sensor, composed of six coupled spiral resonators, exhibits distinct simulated spectral responses under single-gas conditions for hazardous gases relevant to mine safety, including NO \(_2\) , CH \(_4\) , CO, H \(_2\) S, and SO \(_2\) . In its initial configuration, it achieves a sensitivity of 0.947 THz/RIU and a figure of merit of 16.33 RIU \(⁻¹\) . To further enhance performance, a deep neural network surrogate model is developed from WCIP-generated data for rapid prediction and parametric optimization. The model shows high predictive fidelity, with low error metrics and an \(R^2\) value close to 0.998, outperforming several conventional machine learning methods. DNN-guided optimization increases the sensitivity to 1.81 THz/RIU for SO \(_2\) , while preserving excellent agreement with full-wave simulations. A conceptual 6G-enabled IoT architecture is finally introduced to support real-time transmission, edge–cloud intelligence, anomaly detection, and decision support for mine safety. The proposed framework establishes an integrated route toward intelligent, high-performance, and scalable gas monitoring in next-generation smart mining systems.