The Mine Sustainability Modeling Research Group of the Department of Mining and Explosives Engineering at Missouri S&T, as part of the NIOSH project titled “Research, Technological Innovations and Human Factors for Effective Miner Self-Escape from Underground Mine Emergencies” has been developing autonomous capabilities for a quadruped robot. The research, which started in 2021, kicked off with the purchase of the Boston Dynamics’ Spot Dog Robot. The aim was to use the robot to assist miners in self-escape during underground emergencies.
In the first phase of the research, the team conducted a semi-structured focus group where miners used the robot without any prior training. After that, the team asked miners three broad questions: how would they use such a robot on a daily basis? What would they use the robot for during an emergency? How easy is the robot to use? Figure 1 shows the mission type – frequency from the focus group.
Upon analyzing the results from the focus group mission type–frequency, the research team realized that “go find” missions were dominant among all the missions. The researchers now assessed what a “go find” mission entails and settled on two main things, viz, autonomous navigation, and object detection. To develop the autonomous navigation capabilities of the robot, the team needed to integrate a LIDAR sensor to provided better perception in the low-light environment of a mine. The team used the map generated by the LIDAR sensor to develop a path planning algorithm for the robot to navigate underground. During real time deployment, the robot will also employ the LIDAR for localization of its environment and map matching to follow the planned path. To ensure that the robot safely reaches the target location, the team developed control algorithms to control the robot’s movements within the underground mine environment. Figure 2 shows a planned path for navigation in an underground mine environment.
To test the robot’s ability to navigate autonomously using the localization, path planning, and control algorithms, the team conducted several demonstrations runs at the Missouri S&T Experimental Mine (see video below for example). The team also successfully demonstrated the autonomous capability during the second project update meeting at the Missouri S&T Experimental Mine. The footage is shown in the inserted video.
The research team included Dr. Ajay Gurumadaiah, postdoctoral fellow who led the autonomous navigation development, and Cyrus Addy, PhD student who led the object detection work. Both researchers worked together on field trials. Drs. Kwame Awuah-Offei and Siddhardh Nadendla provided guidance and mentoring. When you mention names, link to their profiles on our website.