Enabling Miner Self Escape

Signaling to Enable Miner Self Escape

Background and Challenges

Supporting Agencies

ISC

Contributors

NSF

Thrust 1: Effect of Text Message Alerts on Miners’ Evacuation Decisions

This work evaluates the effect of increasing level of detail in emergency alerts on underground miners’ emergency evacuation decisions. An offline survey was administered to underground miners, which provided miners simulated alerts with ever-increasing levels of information about fire, cave in, and explosion emergency scenarios. The simulated evacuation exercise evaluated differences in evacuation decisions due to the channel of communication (text message or through a co-worker). The results indicate that the content of the alert messages has a significant effect on miners’ decisions with more miners choosing to evacuate when more detail is provided about the urgency of the situation. The results also indicate that the evacuation decisions of the participants are related to their mining experience, marital status, age, income, and use of technology. Depending on the scenario, 51–70% of the miners chose not to shelter in place or go to the refuge chamber as instructed in the simulation. This tendency to ignore guidance is possibly due to a lack of clarity in the message sent to them or the poorly crafted messages that do not account for how the miners will account for risk under uncertainty. It appears that the miners are acting in a loss-averse manner given the uncertainty, as per prospect theory. This study also concludes that, overall, the channel of communication does not significantly affect the miners’ evacuation decisions. This may bode well for using text message alerts in mine emergency management. This work presents a foundation for developing effective text messages for guiding miners in emergencies by building off existing research on short-form emergency warnings and applying the principles to the mining context to provide evidence for new emergency alert technologies mandated new underground coal mine safety regulations. Future research should evaluate miners’ decision-making using prospect theory to account for miners’ perception of risk under uncertainty, so that persuasive text messages can be designed to elicit safe evacuation decisions.

References
  • A. U. Rehman, T. Lyche, K. Awuah-Offei, and V. S. S. Nadendla, “Effect of Text Message Alerts on Miners’ Evacuation Decisions,” Safety Science, vol. 130, Oct 2020

Thrust 2: YOLO-Based Miner Detection using Thermal Imaging for Dark Underground Mines

Surface robots in modern underground mine rescue operations suffer from several limitations such as injured person detection in close spaces, and locomotion difficulties in enabling a prompt self-rescue. Therefore, the possibility of designing and deploying in-mine robots to expedite miner self-rescue, can have an impact on miner rescue during emergencies with lead to reduced fatalities. These in-mine robots for miner self-rescue can be envisioned to carry out diverse tasks such as object detection, autonomous navigation and payload delivery. Specifically, this paper investigates challenges in the design of object detection algorithms for in-mine robots using thermal images, especially to detect people in real-time, in low-light conditions. A total of 125 thermal images were collected in the Missouri S&T Experimental Mine with the help of student volunteers using the FLIR TG 297 infrared camera, which were pre-processed and split into training and validation datasets with 100 and 25 images respectively. Three state-of-the-art, pre-trained real-time object detection models, namely YOLOv5, YOLO-FIRI, and YOLOv8, were considered and re-trained using transfer learning techniques on the training dataset for 50 epochs. After training, the accuracies showed that YOLOv8 was superior to the other two algorithms i.e., YOLOv5 and YOLO-FIRI. On the validation dataset, the re-trained YOLOv8 outperforms the re-trained versions of both YOLOv5, and YOLO-FIRI.

References
  • C. Addy, V. S. S. Nadendla, and K. Awuah-Offei, “YOLO-based Miner Detection using Thermal Imaging for Dark Underground Mines,” Under Preparation for TBD, 2024.