According to the U.S. Mine Safety and Health Administration, 42 miners died in fiscal year 2023, the highest number since 2014. While training, risk assessment, and emergency preparedness and response planning have improved overall mine safety, common hazards such as cave-ins and fires remain.

In collaboration with researchers at the Indian Institute of Technology (IIT) and Indian School of Mines (ISM), Computer Science and Engineering Assistant Professor Sanjib Sur recently began a project to design and deploy an emerging imaging and networking technology to improve safety and overcome hazardous conditions of underground mines.

Sur’s three-year research project began on May 1 and is funded by $827,736 from the National Science Foundation (NSF) and India’s Department of Science and Technology (DST). 

Sur intends to utilize millimeter wave, which is already used as part of the 5G global wireless standard, to help perceive and update a mine environment in real time. Since millimeter wave frequencies provide large bandwidth, the project’s goal is to also enable high throughput communication between miners and autonomous devices for real time updates and decision making. 

“For example, with millimeter wave technology, if there’s a crack or fault on a mine wall, you would get a real time update,” Sur says. “Millimeter wave has been tested with prototype systems in lab but have not been deployed widely. Bringing this transformative technology to a vital sector like mining is important.” 

While more autonomous systems will be deployed inside mines in the future, modern technology will still be needed for miners to safely navigate inside low light or dark conditions where obstructions may exist. Moreover, even if humans are completely removed from mines in the future, autonomous systems will still need millimeter wave technology when traditional sensors like cameras or Light Detection and Ranging (LiDAR) fail in low light or with dust and debris in the air. 

Sur’s team will develop the technology framework, known as MineSense. This includes the millimeter wave hardware component in the core and software tools. The software tools consist of sophisticated signal processing and customized machine learning models to map the environment, provide real time update and decisions, and enable high-throughput communications. 

The project will consist of three tasks. The first will be creating real time 3D maps for the inside of mines and to determine which parts require more machine deployment. Sur plans to use millimeter wave in the form of a small chip installed on a miner’s helmet. As the miners walk and move their head, the chip will constantly transmit and receive millimeter wave signals, while an algorithm will reconstruct the environment in real time. 

“For 3D mapping and communication, the only option with existing systems is to install a separate network and access point and devices, which can be costly to maintain. But millimeter wave enables dual networking and sensing framework to understand the environment and communicate data within the infrastructure,” Sur says. 

Sur admits that some may question why a real time update is needed since the mine infrastructure is already known. But walls may partially move or collapse, new areas may be excavated, and faults and cracks may leak, which a static map would be unable to discover. A static map will be augmented to create updated real time vital signs of the mine. Sur hopes that as 5G networks become more abundant, hardware costs will decrease to make this a cost-effective and scalable solution. 

Enabling a low latency communication (networking) infrastructure with miners and machines is the second task. Sur’s team aims to deploy routers and access points inside specific locations of the mine to establish communication across different devices as miners are working. As part of the task, testing will be performed at different millimeter wave frequencies. For example, if a miner is walking underground with their device, the information from their device needs to be uploaded and merged with the other miner’s information and combine it with the static map itself. That requires high speed and high throughput communication.

A comprehensive evaluation is the third and most critical task. Sur wants to bring his work into a mining simulator, which is located at ISM and includes rails and cutting machines. It also has walls lined up with mining materials such as coal and minerals. 

“We want to leverage the infrastructure in the simulator and eventually put these systems inside a real mine in India to evaluate how well we are doing beyond the lab,” Sur says. “If we demonstrate the practicality of this technology, we can create more collaborations with the mining industry.”

Sur’s collaboration with institutions in India emerged through an NSF-DST solicitation. Part of the collaboration came from Sur’s previous work on using millimeter wave with autonomous vehicles to safely navigate hazards on roads at night. 

“While discussing with our collaborators in India, the topic came up because mining accidents in India are a big concern. We saw an opportunity to provide technological solutions with substantial social benefits,” Sur says. “We believe this collaboration will significantly enhance the mining industry in both countries.”

IIT Associate Professor Sandip Chakraborty, who is Sur’s partner on the project, says that the mining industry has unique challenges with communication and sensing. But if millimeter wave technology is successful, the same platform can be used to disseminate useful information among miners, while also sensing dangers.

“Imagine a scenario when the mining worker is talking over a communication device, and it senses a possible landslide from behind. The device can immediately warn the worker and help them to free the space with near-zero delay,” Chakraborty says. “I am particularly looking forward to developing a technological solution by utilizing the expertise from the three institutes in unique directions.” 

Sur says the most exciting part of the project is bringing modern technology to a vital sector like underground mining for improving safety and efficiency.

“We feel that millimeter wave system will be like a Swiss army knife in the future,” Sur says. “If designed well with sophisticated signal processing and deep learning tools, it can be applied in many other sectors.”