Leveraging AI and satellite technologies for autonomous operations

It is becoming increasingly apparent that AI, satellite networks and linked technologies, such as automation, are vital to getting more discoveries. Photo by Scharfsinn86 via Getty Images

The mining industry is in the process of an AI and automation upgrade, catalysed by the energy transition and an increased demand for lithium, copper, rare earths and other critical minerals crucial to its success.

Mining companies are battling to supply more of these minerals used in renewable technologies such as electric vehicles, wind turbines and energy grids.

However, Jef Caers, founder of Mineral-X, an industrial affiliates programme at Stanford University’s School of Sustainability, focused on technological innovations for the critical mineral supply chain, says: “We are putting more money into exploration and getting fewer discoveries.”

It is becoming increasingly apparent that AI, satellite networks and linked technologies, such as automation, are vital to solving this problem.

From his vantage point in the space sector, Dr Craig Brown, director of investment for the UK Space Agency, has witnessed a notable increase in the number of companies adopting these new technologies. He notes that approximately 40% of applications for the agency’s most recent development programmes incorporate AI on some level, and many would have direct application in the mining sector.

For Caers, self-driving vehicles are a clear example of how automation is enhancing mining operations.

“We plan thinking we know the answer. We know the ore body, we know the earth, we know everything.

“If you’re mining the wrong way, the robots are going to fix that,” he explains.

Analysis from MINE’s parent company GlobalData reveals a rapid rise in the industry’s use of autonomous, autonomous-ready or tele-remote equipment in recent years, which now represents over 4% of all key mining equipment.

In July, GlobalData identified 3,832 autonomous haul trucks operating on surface mines across the globe. This figure includes both those that are running autonomously and are autonomous-ready.

Caers’ programme has partnered with US-based mine exploration company KoBold Metals, for a project that has optimised AI and automation technologies in Zambia to help make what they describe as the largest copper discovery in a decade. Tech-focused and backed by the likes of billionaire businessmen Bill Gates and Jeff Bezos, KoBold positions itself as a challenger to traditional mining companies.

Autonomous robots assess the world around them, whether that be via sensors, global positioning systems, advanced software or robust wireless networks, to carry out mining operations with greater precision than humans. This precision also results in limited waste, while the lack of human involvement allows companies to conduct operations remotely.

Satellite technology is integral to the remote element of automation, supporting autonomous equipment such as trucks and drills by providing robust connectivity, precise positioning via global positioning systems, for example, and real-time data for the planning and monitoring of a mine.

It underpins a digitalised mining operation, otherwise known as a smart mine, helping to make operations more efficient, safer, and environmentally compliant.

“You need to communicate, or you need to take imagery. You need to navigate in areas where there isn’t very good ground infrastructure, where you’re in remote regions. It could be out in the desert, it could be forest regions, it could be at the ocean,” Brown explains.

“This is where space really comes to the fore because there’s no other way of doing this stuff other than by satellite.”

Companies are already using, for example, analysis techniques on the behaviour of tired workers in the field, such as drivers using heavy equipment. Information is processed by a camera and a microphone for signs of fatigue in their voice or body language, says Brown.

As the drivers often work in remote regions, the data is sent to the company headquarters (HQ) with the help of a satellite. HQ can then understand if the drivers are doing something which could endanger their lives or others’.

“Sometimes satellite isn’t the whole answer, but it’s an important part of the value chain to make sure service is delivered where you need it,when you need it,” says Brown. 

Autonomous mining technologies can often work in tandem with AI, including for planning. AI technologies such as generative models, digital twins and edge computing are now integrated into mine planning systems to help monitor, optimise and scale operations. 

Mineral-X and KoBold’s partnership led to them developing an AI algorithm for drillhole planning that resulted in the rare ultra high-grade discovery in Zambia.

“While human experts are involved in planning, AI can calculate the risks of human plans, which can help mining companies mitigate better against unexpected outcomes,” explains Caers.

“In Zambia, you are mining a very thin layer – the ore shale. If you can’t follow it, it’s not like a big ore body that you just go in and break down. If you follow it incorrectly, let’s say the layer goes one way, but you’re mining the other way. That’s a huge cost.”

It’s these details that he believes are important to plan for, and with AI introduced into these planning systems, “you get better results”. 

For KoBold Metals, following a discovery in 2023, it has set its sights on opening a mine at the site in seven or eight years – half the normal time. Caers says that if KoBold succeed in sticking to this timeline, the reason would be AI.

While Mineral-X and KoBold are helping to push AI and automation to the fore in mining operations, Caers raises a second problem: mining companies are mostly unaware of these technologies at present. Self-driving technologies as we know them are still in their infancy, he explains.

“Now these technologies are maturing in the tech sector; they are going to make their way through into the mining world.”

However, there are challenges to this progression, which Brown has witnessed in the space domain: “One of the challenging trends I think we’ve got within the space domain is that the level of ambition very often outstrips the size of our budgets as governments.”

According to Brown, the solution to the problem is “catalysing more and more private capital into the sector”. The Space Agency helps with this by supporting new investors.

As for the mining industry, Caers believes it needs to rethink historic processes and planning deterministically. He says quantifying uncertainty about the subsurface is imperative.

While he notes that quantifying uncertainty is challenging, he suggests that the AI community has figured out that, with the technology, when you make a prediction, you don’t just get a prediction; you also get a probability of you being right about the prediction, which helps in making better decisions.

For Brown, any company that wants a commercial advantage and is keen on innovation and staying ahead of the curve “should consider the new services that are coming through some of the technology [the space domain] supports”.

As for where AI and autonomous technologies are headed, Brown views the collaboration between the space domain and the mining industry as a two-way street. He believes the techniques and capabilities within the mining sector could complement AI-assisted autonomy in the space domain to aid better projects in space.

Brown comments: “There are plans globally around the exploration of the Moon and Mars, and if we are to secure a permanent off-world presence on either or both, we’re going to have to find ways to use the resources that are available on those kinds of non-terrestrial bodies.”

He says we are already seeing some interesting projects run by companies in the US that are looking at this, including those by Blue Origin, founded by Jeff Bezos.

He adds that certain developments in these projects will not happen without figuring out the extraction of material from the lunar surface, and that’s where mining technologies will be at the forefront of the future.