IoT-Based Predictive Safety System for Coal Mines Using Deep Learning Techniques

In order to proactively identify and reduce possible risks, this study presents a cutting-edge IoT-based smart safety system for coal mines that integrates machine learning and real-time environmental monitoring. This system simulates a variety of mining conditions, including as temperature, humidity, CO<inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</inf> levels, and ventilation rates, utilizing synthetic sensor data produced using statistical distributions, in contrast to conventional systems that rely on simple thresholding or algorithms. To categorize environmental conditions as either safe or dangerous, this data is processed using a deep neural network model constructed with Tensor Flow. To ensure a thorough assessment of the model's predictive skills, it is trained and validated using a variety of performance indicators, such as confusion matrix, F1 score, recall, accuracy, and precision. The system is further optimized using performance visualization tools, such as confusion matrix heat maps and accuracy/loss graphs. After optimization, the model is included into the mine's Internet of Things infrastructure to deliver hazard alerts in real time, improving the mine's ability to respond to dangerous situations. By seeing hazards before they become serious, this proactive approach to hazard identification enhances mine safety and represents a major improvement over traditional safety system.