Data analytics can help mining companies ensure the safety, sustainability and profitability of their operations, but only if the technology is applied in the right way and at the right time. We speak to mining experts from AMC about the optimal use of data and how to make the most the competitive advantage it brings.
In November 2020, Shenbao Energy and Aerospace Heavy Industry, a Chinese mining truck manufacturer and technology company, successfully tested five 5G autonomous trucks in extremely cold conditions at Shenbao’s Baorixile coal mine in Inner Mongolia.
A little earlier the same year, in Papua New Guinea, mineral exploration company Freeport Resources, deployed Minerva’s Driver AI solution at its Star Mountains project to enhance surface and sub-surface exploration of the property.
Meanwhile, in the far east of Russia, diamond mining company Alrosa installed an automated wireless monitoring system at its Aikhal Division as well as new wearable devices for employees.
What all these projects, along with hundreds of other examples across all size of company and mining operation have in common is their reliance on the generation, collection, organising and analysis of huge quantities of data.
A further illustration of the data-driven, digital revolution underway in the mining industry, comes from GlobalData’s prediction that mining firms’ spending on AI platforms will reach $218m by 2024. This is up from $76m in 2019, and represents annual growth of 23.4%.
And analysis of GlobalData’s unique mining jobs database shows that recruitment patterns across the industry offer more evidence of this trend with shows that jobs linked to data being among the most heavily advertised in the first quarter of 2021.
The data revolution in mining
The exponential growth in data generated by modern life is a universal truth. And it goes as much for the mining industry as any other. Mining has embraced the digital and data revolution, with the central role of data across all areas of operations a well-established phenomenon.
From improving daily operations, via real-time tracking of people to enhance safety, to monitoring machinery and predicting when maintenance will be needed; or using GPS to accurately track locations and ensure shovels are digging when and where they should be (to the nearest centimetre), data underpins all mining operations and has an impact on everything.
Despite the investment, challenges remain
Standing in the way of maximising an organisation’s return on data are ensuring that the right data is collected, its quality and reliability is assured, and clear strategic thinking is employed to get the most out of data. Many decision makers are still not fully exploiting all the data they collect.
Andrew Hall is a director and executive lead for advisory at AMC Consultants. He sympathises with the struggles miners face in getting a handle on all this data, highlighting the rapid cultural change that has occurred from the days, not long ago, when mines generated little data, to now having seemingly endless quantities available on all aspects of an operation.
“The industry’s greatest opportunity is how best to use that information,” he observes. “There’s a lot of it, and it’s easy to lose focus on the real value drivers. Sometimes analytics can drive you in the wrong direction if you don’t maintain a holistic perspective. The learning curve for a mining project never finishes,” he says.
Hall adds that this is an area where the value of deep industry-specific understanding comes into its own. Interpreting data is something that can’t be done well without industry-specific context.
This matters because if data isn’t collected, stored and managed efficiently and if the right tools or metrics are not used to analyse it, they may not only fail to improve things, but they may miss the optimum plan or strategy and end up failing to realise the full potential value from a deposit, or worse, expose their mineral asset to greater risk.
Damian Peachey, advisory lead and principal engineer at AMC, agrees. He points out that while most operators use data now, they use it in a wide range of formats, from basic spreadsheets and PDFs to sophisticated platforms. And not enough are getting the best from what’s gathered.
“Some try collect so much raw data they risk ending up down a rabbit hole,” he observes. He suggests, it would be more effective to analyse the key data necessary to focus on high-impact areas.
It’s not what you’ve got, but the way you use it
Good processes in data management and using data effectively are vital to realising value from technology. Having put time and effort into gathering data, it makes sense to also make sure the data is clean and meaningful.
Metadata explains under what conditions the data was measured and is vital to understanding the quality and reliability of the data. Collection of metadata needs to happen from the beginning of the process – when a deposit is first discovered – through to the conclusion when the product leaves the plant.
Data offers insight and opportunity in predictive geometallurgy – the business of using geological, metallurgical, and mining, data to predict the way ore will perform when it goes to the processing plant.
As Ian Lipton, principal geologist at AMC, explains, “How much energy might the ore consume, how much reagent will be necessary to extract the mineral of interest, how much of the mineral are you going to recover; all these factors flow through to the cost of recovering the mineral you’re interested in,” he explains. “So, the better you are able to characterize the ore, the more opportunity you have to design and operate in ways that maximise your margins.”
Miners are beholden to the vagaries of geology. “Ore bodies are intrinsically variable on every scale right down to the millimetre, and only a tiny fraction of an orebody can be sampled before it is mined” says Lipton. “At the feasibility stage, the data is sparse, which makes development of predictive relationships and forecasts difficult. But the application of modern data science and machine learning are enabling step-change improvements in predictions.”
Once in production, a well-instrumented mine produces vast quantities of data; so-called big data. “Some operators are becoming really good at leveraging the information generated in their mines to optimise operating practices,” adds Lipton. “They are joining front-end data from the mineral resource model and back end data from the process plant to continuously improve predictions of the way the ore is going to behave. The more accurately you can predict the ore behaviour, the more opportunity you have to be prepared and manage the natural variability of the ore by informed blending and smart scheduling.”
Successful operations share data across teams
Companies that are successfully using data to make informed decisions to improve operations do so by collaborating throughout the organisation – from fitters, engineers and geologists all the way to the CEO, says Jayson Tolley, advisory lead underground at AMC.
“It’s about understanding the reality of every layer of the business and ensuring the data is shared so there is one source of truth. There’s a link between everything. It’s all interrelated, from the deposit all the way through to final product.”
Lipton agrees. He reckons the best mines operating today are the ones that understand their data and share it across the organisation. In addition, they know when they need to bring in expert consultancy or change tack, switching off the existing plan to something likely to bring better results.
“Many of the big mining companies have invested heavily in data collection and then use specialist data science to identify the root cause of problems that are not identifiable by simply looking at spreadsheets,” he explains. “Data science has provided insights that weren’t achievable by conventional means.”
Don’t look for a silver bullet
Still, data is no silver bullet, says Hall. He reiterates that strong collaboration within and between disciplines is vital, with everyone looking at the whole mining value chain, not just their small area.
Peachey agrees. Noting that without experience and that deep domain understanding, data itself won’t help: “ A proper understanding of context is crucial for correctly identifying the root causes of a problem. Then, armed with that knowledge you can make better informed, practical decisions that provide sustainable outcomes.”
Or bring in the experts who can, as Lipton explains it, help operators make sense of their data, undertake the right analytics, and avoid the pitfalls. “Increasingly, the thread that runs through every aspect of a mining operation is data. If you understand the data and choose the right methods to extract information and insights, then information from each part of the mining process can be used to inform and improve the performance of the whole. That’s how smart data adds value to a mine.”
By clicking the “Submit” button above, you accept
and acknowledge that your data will be used