Applications of AI in the mining sector

In July 2024, Barrick Gold partnered with Australian space technology firm Fleet Space Technologies to enhance copper exploration at the Reko Diq site in Pakistan, which it claims is one of the world’s largest undeveloped copper-gold mines. Barrick holds a 50% stake in the project and has committed a $10 billion investment, with the remaining ownership likely to be evenly divided between the Pakistani and Balochistan governments.  

Barrick will use Fleet’s ExoSphere system, which uses a low-Earth orbit satellite network, smart seismic sensors with edge computing, and data analytics to generate 3D sub-surface maps and identify potential copper mining zones at Reko Diq.  

Fleet claims ExoSphere discovers critical resources up to 100 times faster than traditional methods while minimizing the environmental impact by reducing the need for explosives, noisy machines, Source: Fleet Space Technologies and extensive drilling. This technology will allow Barrick to achieve more accurate resource identification, speeding up decision-making and enhancing exploration outcomes. Barrick plans to start production at Reko Diq by 2028.

In July 2023, Almonty Korea Tungsten Corp (AKTC), the South Korean subsidiary of Almonty Industries, partnered with Korea Telecom (KT) to launch the Mine Safety DX solution at the Sangdong mine, one of the world’s largest tungsten mines.   

The Mine Safety DX system integrates wearable tech, AI-driven monitoring, and access and location management to enhance mine safety. Smart bands and helmets equipped with smart tags collect real-time biometric data from miners while they work. AI algorithms analyze this biometric data to detect signs of fatigue and physical distress, enabling proactive interventions to remove miners from hazardous situations before accidents occur. The tags also track miners’ locations and issue notifications when they enter dangerous areas, raising awareness of risks and ensuring backroom staff remain ready on standby. This real-time tracking enables swift responses during emergencies, with quick location identification crucial for saving lives.  

The system runs on a 4G network using underground cabling rather than traditional routers. This setup ensures connectivity during cave-ins or power outages, allowing communication through regular phones if needed. Almonty has pledged to make this technology open source for all South Korean mines.   

Almonty aims to integrate autonomous equipment into the system within three years. This equipment will work with existing devices to identify and respond to potential dangers, similar to collision avoidance technologies. Almonty has introduced several other innovations to improve mine safety in recent years, with the Mining Safety DX system following the company’s implementation of a lighting system that mimics daylight underground to reduce fatigue, and an automated plant that can be operated by only four people.

In July 2024, Fortescue and Swiss original equipment manufacturer (OEM) Liebherr Mining announced a partnership to develop what they claim will be the world’s first fully integrated autonomous haulage system (AHS) for zero-emission vehicles. The AHS will be deployed at Fortescue’s iron ore operations in Western Australia’s Pilbara region, supporting the company’s goal to achieve net-zero Scope 1 and 2 emissions across its Australian iron ore operations by 2030.  

The new AHS integrates a fleet management system (FMS), an onboard autonomy kit for Liebherr’s T 264 trucks, and a high-precision machine guidance system for loading equipment. The FMS is designed to optimize fleet utilization, boost productivity, and enhance fuel and energy efficiency. The AHS will enable the control and coordination of mixed fleets, including ultra-class mining trucks, road trains, and light vehicles, and will be available to customers worldwide. The AHS and a fleet of four T 264 autonomous trucks are currently being tested and validated at Fortescue’s Christmas Creek mine site.  

Fortescue previously completed one of the industry’s largest fleet conversions to autonomous vehicles in 2020, transitioning 108 haul trucks. The company has said its autonomous fleets have achieved a 30% increase in productivity compared to previous manned operations.   

The latest collaboration builds on the existing alliance between Liebherr and Fortescue, which started in June 2022. The two companies aim to create a zero-emissions haulage solution using Liebherr’s OEM expertise and Fortescue’s advanced fuel cell, battery, and battery-electric technology. In August 2024, the first hydrogen-powered T 264 haul truck prototype, developed by both companies, arrived at Fortescue’s Christmas Creek mine for real-world testing.

In August 2023, US Steel announced a partnership with Google Cloud to deploy MineMind, a generative AI-powered application designed to streamline equipment maintenance, at its Minnesota iron ore operations. US Steel hopes MineMind will drive efficiency and cut costs.  

MineMind allows US Steel maintenance teams to consult AI for immediate answers to technical questions. US Steel technicians can query the system for issues like troubleshooting hydraulic system malfunctions, calibrating complex sensors, or replacing worn conveyor belts. MineMind generates detailed summaries by retrieving relevant standard operating procedures and providing complex diagrams, eliminating the need to sift through extensive technical manuals. MineMind also provides a validity score and source references for any information provided, ensuring technicians can rely on the accuracy and trustworthiness of the guidance they receive.  

MineMind’s implementation began in September 2023, initially focusing on A60 haul trucks at US Steel’s Minnesota Ore Operations. US Steel suggests the tool reduces the time required to complete work orders by approximately 20%, enhancing productivity and reducing maintenance costs. US Steel plans to work with Google Cloud to develop generative AI applications that address high-impact areas across the entire US Steel value chain.