Collabs with academia, tech firm, with funding from government for future safe, smart vehicles.

With 150 years of experience in all things mining, Sweden-based Epiroc has unveiled its next generation in mineral extraction, which will leverage the power of artificial intelligence to make its mining vehicles more autonomous through a new collaborative project called the Realistic Navigation Assistance and Machine Control, or ReNAM.

Funded by Swedish government innovation agency Vinnova and spearheaded by Luleå University of Technology and its spin-out, Algoryx Simulation AB, this all-Swede project aims to leverage simulations and cutting-edge research on Epiroc's underground mining machines to achieve a higher level of autonomy and safety.

"We are doing so much more than autonomy, essentially, we are accelerating equipment to a new level of technology," said George Nikolakopoulos, professor of robotics and AI at Luleå University of Technology. "The overall quality of Swedish mines, efficiency and safety will be unparalleled."

With the global mining industry facing increasing demand for metallic commodities, most newly discovered ore deposits have been located far below the ground, at depths of more than 300 meters.

According to Luleå University, there are plans to mine at a depth of more than 2,000 meters in Sweden.

Mining over a mile below the ground is more than just challenging; for every meter, risks and costs increase exponentially, especially if the cost is a safe working environment.

This is where autonomous vehicles can reduce these human risks and ultimately make mining cheaper – by removing people from them – consequently cutting out much of the infrastructure that facilitates safety in the first place.

"One of the most important technology shifts in the mining industry involves autonomation solutions enabling operators to work remotely from safe and more accessible locations," said Oskar Lundberg, global innovation manager at Epiroc Underground. "This requires new systems which give the machines 'eyes and ears' to assist the operator. The development of these systems is complex and can only be successful through close collaboration between academia, small technology firms and machine suppliers."

In its initial stages, the project will first involve creating realistic digital twins of mining machines and their surroundings in a physics-based simulation environment. Essentially replicating scenarios and environments for future remote operators, these virtual mining machines will be equipped with sensors and AI capabilities to safely and reliably perform tasks in this simulated environment.

Once this portion is completed, the team will transfer this dataset and control systems to Epiroc's actual mining machines to begin testing performance in the mines.

"Through realistic simulations, where the AI can be trained on the whole system (the machine) with its environment and real forces, contacts and behaviours, with safety in mind, the outcome will be hugely superior to most projects we see in the world," said Urban Wikman, business director at Algoryx. "To be successful, collaboration between universities, high-tech and industrial companies is extremely important."

Kicking off this year, ReNAM will run to the end of 2025. Ultimately, this project aims to develop and validate two use cases that lay a foundation for future robotics research and contribute to Epiroc's future development of commercial products with fully developed machine autonomy.

"Through this partnership, we are contributing to a new generation of machines with improved autonomy capabilities," said Åsa Gabrielsson, vice president of research and development at Epiroc Underground. "This is an exciting addition to the on-going development of our automation offering, which helps to further improve safety and productivity for our customers."