Autonomous haul trucks have become a familiar part of the mining industry over the past decade, just look at the iron ore sector in the Pilbara region of Western Australia.
Mining companies have also started to introduce autonomous drill rigs and even heavy-haul, long distance trains.
While autonomous equipment has emerged in some areas of mining, minerals processing remains a sector where uptake of the technology has been limited.
That’s not to say it isn’t possible. According to Metso chief digital officer Jani Puroranta, there are great benefits to be gained by aiming for what he calls “full autonomy”.
Puroranta says full plant autonomy requires an interplay of advanced process control augmented with new sensors and analysers; highly skilled operators aided by artificial intelligence (AI) and supported by remote domain experts; hi-fidelity, dynamic, real-time process simulation (digital twin); and 100 per cent predictive maintenance, with no unplanned shutdowns.
“I think it’s a big change management issue,” Puroranta tells Australian Mining at the International Resources and Mining Conference (IMARC).
“You can add software to a plant, but it is very hard to educate people and get them to change. If you don’t educate people and centrally change their behaviour it is not going to stick.”
Puroranta has led Metso’s development of a five-level framework that aims to guide mining companies towards fully autonomous plant by 2050.
The framework starts with regulatory controls and progresses through advanced process control, intelligent equipment, analytics and AI, and full autopilot (autonomy).
Apart from a few rare exceptions, usually sites operated by major global miners, Puroranta believes the industry’s maturity is mostly at level one.
The majority of companies have either not properly executed level one, he adds, or have lost their way by not taking an orderly approach when moving through the steps.
“If they are at a basic level and struggling with automation, we can add sensors – we could do that to a crusher for example,” Puroranta says.
“I think many companies are struggling to maintain level one. There is a lot of work involved in maintaining this level, but you can’t stop there because that is the most basic level.”
Metso works with its mining customers interested in realising the benefits of fully autonomous plant to analyse which level of the framework they are at and which direction to move in next.
Puroranta says it is vital for Metso to understand where an operator’s “pain points” are, whether it is in crushing, screening, grinding, flotation or another part of processing.
“It is a matter of getting a better grip of the process. Is advanced process control the way to go? Or should we add intelligent equipment, or should we add AI? You can make these decisions when you know which maturity level you are at,” Puroranta says.
“Then we can start figuring out the solution. It is usually a combination of things – it is not just one thing. You might want to try many things at the same time.”
In contrast to the perception that autonomous plant means fewer people are required to manage mining operations, Puroranta says human workers have never been more important.
He believes how mining companies integrate people into the projects will determine how successful they will be.
“That is the biggest part that you need to address and sometimes people don’t know that is what it is all about,” Puroranta says.
“They are scared about their jobs, but it doesn’t take away their job – it makes it more interesting for the person.”
Automation projects at processing plants are also best suited to brownfields operations, another notable difference when compared with projects for underground operations or haul trucks.
Metso is able to identify the “pain points” at a brownfields operation by working with its customer, making it easier to know where to apply the framework and focus investment.
At a greenfields site, on the other hand, there is a risk to overinvest in automation and complicate the broader operation by moving deep into the framework when there’s no need to.
You can add software to a plant, but it is very hard to educate people and get them to change.
Metso highlighted two new pieces of intelligent equipment at IMARC that have been specifically designed to facilitate the industry’s gradual move towards autonomous processing plants.
Its new gyratory and cone crushers, and the revolutionary VPX filter for dewatering of tailings are powered by the Metso Foresight data system and aim to enhance how operations analyse, adapt and anticipate activities.
The company is giving companies opportunities to move as far as levels three and four with these new solutions, making autonomous processing a reality for the more mature companies.
Despite this potential, Puroranta backs Metso’s conservative view urging the industry to develop a vision for full processing autonomy by 2050.
“Advanced process control is level two and that has been around for more than 20 years,” Puroranta says. “It has been rolled out since at least the 1980s and it is not mainstream in the industry today – there are still a lot of companies thinking about it.
“That has taken more than 30 years to implement so I think another 30 years is a more realistic timeframe (for full autonomy), but I would hope in the next 10 years we have gone a fair way.”