Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Secure and Privacy-Preserving Federated Learning for Next-Generation Underground Mine Safety

Mohamed Elmahallawy, Sanjay Madria, Samuel Frimpong

TL;DR

This paper tackles privacy and security challenges in deploying machine learning for underground mine safety by introducing FedMining, a privacy-preserving federated learning framework. FedMining employs a decentralized functional encryption-based secure aggregation (IPFE-derived) and a balancing aggregation scheme to mitigate non-IID data bias and model staleness, enabling fast convergence without exposing local data or models. The approach demonstrates high hazard-detection accuracy (average mAP around 0.83 across classes) and significantly reduced communication and computation overhead compared to Paillier-based methods, achieving up to ~27x lower communication and ~14x lower encryption-time cost. Real-world mining datasets (DsLMF+) and standard benchmarks (CIFAR-10/100) validate robustness to non-IID settings and show rapid convergence, making FedMining practical for real-time underground safety monitoring while preserving data confidentiality.

Abstract

Underground mining operations depend on sensor networks to monitor critical parameters such as temperature, gas concentration, and miner movement, enabling timely hazard detection and safety decisions. However, transmitting raw sensor data to a centralized server for machine learning (ML) model training raises serious privacy and security concerns. Federated Learning (FL) offers a promising alternative by enabling decentralized model training without exposing sensitive local data. Yet, applying FL in underground mining presents unique challenges: (i) Adversaries may eavesdrop on shared model updates to launch model inversion or membership inference attacks, compromising data privacy and operational safety; (ii) Non-IID data distributions across mines and sensor noise can hinder model convergence. To address these issues, we propose FedMining--a privacy-preserving FL framework tailored for underground mining. FedMining introduces two core innovations: (1) a Decentralized Functional Encryption (DFE) scheme that keeps local models encrypted, thwarting unauthorized access and inference attacks; and (2) a balancing aggregation mechanism to mitigate data heterogeneity and enhance convergence. Evaluations on real-world mining datasets demonstrate FedMining's ability to safeguard privacy while maintaining high model accuracy and achieving rapid convergence with reduced communication and computation overhead. These advantages make FedMining both secure and practical for real-time underground safety monitoring.

Secure and Privacy-Preserving Federated Learning for Next-Generation Underground Mine Safety

TL;DR

This paper tackles privacy and security challenges in deploying machine learning for underground mine safety by introducing FedMining, a privacy-preserving federated learning framework. FedMining employs a decentralized functional encryption-based secure aggregation (IPFE-derived) and a balancing aggregation scheme to mitigate non-IID data bias and model staleness, enabling fast convergence without exposing local data or models. The approach demonstrates high hazard-detection accuracy (average mAP around 0.83 across classes) and significantly reduced communication and computation overhead compared to Paillier-based methods, achieving up to ~27x lower communication and ~14x lower encryption-time cost. Real-world mining datasets (DsLMF+) and standard benchmarks (CIFAR-10/100) validate robustness to non-IID settings and show rapid convergence, making FedMining practical for real-time underground safety monitoring while preserving data confidentiality.

Abstract

Underground mining operations depend on sensor networks to monitor critical parameters such as temperature, gas concentration, and miner movement, enabling timely hazard detection and safety decisions. However, transmitting raw sensor data to a centralized server for machine learning (ML) model training raises serious privacy and security concerns. Federated Learning (FL) offers a promising alternative by enabling decentralized model training without exposing sensitive local data. Yet, applying FL in underground mining presents unique challenges: (i) Adversaries may eavesdrop on shared model updates to launch model inversion or membership inference attacks, compromising data privacy and operational safety; (ii) Non-IID data distributions across mines and sensor noise can hinder model convergence. To address these issues, we propose FedMining--a privacy-preserving FL framework tailored for underground mining. FedMining introduces two core innovations: (1) a Decentralized Functional Encryption (DFE) scheme that keeps local models encrypted, thwarting unauthorized access and inference attacks; and (2) a balancing aggregation mechanism to mitigate data heterogeneity and enhance convergence. Evaluations on real-world mining datasets demonstrate FedMining's ability to safeguard privacy while maintaining high model accuracy and achieving rapid convergence with reduced communication and computation overhead. These advantages make FedMining both secure and practical for real-time underground safety monitoring.

Paper Structure

This paper contains 20 sections, 5 equations, 4 figures, 4 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Non-security-based hazard prediction
  4. Security-based hazard prediction
  5. FedMining System Model, Threat Model, And Design Objectives
  6. Federated Learning in Underground Mining Operations
  7. Threat Model
  8. Design Objectives
  9. Methodology
  10. FedMining Framework
  11. Secure Aggregation Scheme
  12. Balancing Aggregation Scheme
  13. Performance Evaluation
  14. Experimental setup
  15. Security analysis of FedMining
  16. ...and 5 more sections

Figures (4)

  • Figure 1: An illustrative visualization of the FedMining framework.
  • Figure 2: Evaluating mAP changes during training across various DsLMF+'s classes.
  • Figure 3: Confusion matrix depicting the correspondence between predicted outputs of the global model and ground-truth class labels.
  • Figure 4: Samples from the DsLMF+ dataset used to test the prediction accuracy of the global model.