AI based Safety Helmet for Mining workers using IoT Technology and ARM Cortex-M

Coal mining is one of the most hazardous activities in the world. They frequently encountered unexpected emergencies. The use of the IoT and AI in mining helps improve worker health management and prevent injuries. In this study, a Personal Protective Equipment (helmet) is proposed, which can provid...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:IEEE sensors journal 2023-09, Vol.23 (18), p.1-1
Hauptverfasser: Lalitha, L, Ramya, G, Shunmugathammal, M
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Coal mining is one of the most hazardous activities in the world. They frequently encountered unexpected emergencies. The use of the IoT and AI in mining helps improve worker health management and prevent injuries. In this study, a Personal Protective Equipment (helmet) is proposed, which can provide alert signals to the control center to inform the miner about the risk. With the use of several sensors integrated into the STM32 module, it continuously analyzes ambient conditions (toxic gases, temperature, and humidity), as well as the worker's health conditions, such as heart rate and vibration generated by excavation and blasting, which are subsequently relayed to the control center using a low-energy Bluetooth module. This system also has a panic button that may alert the control unit if there are any dangers to the workers. The DHT11 (Digital Temperature humidity Sensor) can measure the temperature and humidity levels with a degree of accuracy that falls within a range of ± 5%. The MQ135 Sensor, on the other hand, can sense gas concentrations with 85% accuracy. In coal mines, high gas concentrations can cause miners to feel dizzy and disoriented. To address this issue, miners can press a panic button located on their helmets, which alerts the control center staff and speeds up the rescue operations. In addition, a heart rate sensor was integrated with the STM module using the I2C protocol. If the heart rate reading falls below 60 or exceeds 100, it is considered an abnormal condition that requires attention. Furthermore, a machine learning algorithm with a convolutional neural network helps to train the artificial intelligence model to recognize the worker's gestures. Here, four types of gestures were fixed, which helped the workers communicate. These gestures have been labeled GOOD, NOT GOOD, DOING FINE, and EMERGENCY EVACUATION. A receiver API is proposed to visualize the results from various sensors and take appropriate action to safeguard miners.
ISSN:1530-437X
1558-1748
DOI:10.1109/JSEN.2023.3296523