The mining industry is facing a growing array of challenges, from volatile markets – not helped in the least by the two-punch combo of the Covid-19 pandemic and the Russian invasion of Ukraine – and tougher competition, to regulatory compliance and decarbonisation, among many others.

A survey published by Ernst & Young (EY) in 2022 looked at the top priorities for mine operators, with environmental, social and governance (ESG) objectives holding on to its top spot from the year prior. Unsurprisingly, geopolitics has soared to second place amid the ongoing global conflict and uncertainty. Digital innovation, on the other hand, has dropped down the rankings all the way to ninth place – however, this isn’t necessarily a bad thing for its proponents, as the fall comes partly due to operators developing confidence and capabilities in this area, as the report notes. Today, many mining companies are already taking advantage of the significant cost, productivity and safety gains from the implementation of emergent technologies like drones, remote operating centres and autonomous trucks, to name but a few.

Yet, the EY report also notes that the mining sector still has a largely siloed approach to digitalisation, holding back its full potential, despite the encouraging progress in recent years. A more integrated strategy would help mine operators tackle their most pressing challenges, including ESG, productivity and cost efficiency. Technological innovation, and artificial intelligence (AI) in particular, is already benefitting the mining industry in a wide variety of ways, from supporting the discovery of more financially viable mineral deposits to optimising operations. However, there’s one area that’s proving pivotal to providing actionable insights for the mining sector, and that’s predictive analytics.

Deep data dive

A huge amount of data is produced every day across modern mining operations, and this hasn’t escaped the industry’s attention. Indeed, mentions of big data within the filings of mining companies rose 30% between the first and second quarters of 2022, according to GlobalData – up 279% since 2016. However, much of this information is wasted – either unused or poorly implemented for providing insights to mine operators. With the right tools, however, this data has the potential to help reduce unplanned downtime, streamline processes, improve asset performance and boost a mine’s ESG credentials.

Predictive analytics solutions use data to state the probabilities of the possible outcomes in the future. Mining companies can then make use of these probabilities to plan many aspects of their operations. The technology is intended to help maintenance planners, systems engineers, controllers and other mine personnel to make real-time decisions that improve performance, reliability and profitability.

With mining operators facing a future of dwindling resources and the challenges presented by deeper mines, rising energy costs and infrastructure shortages, there has never been more pressure to improve efficiency and cut costs. However, as many within the mining sector would be the first to admit, the industry is often slow to embrace change, its conservative instincts instead preferring to rely on tried and tested methods that have served it well for decades.

For many holdouts against predictive analytics, one of the apparent downsides of the technology is that it can seem intimidating and overly complicated, and its return on investment (ROI) is unclear. However, modern solutions don’t require a data scientist to be at hand to model and configure the application – instead, they’re designed to be easy to use and implement across daily operations and aim to provide quick payback for mine operators. For example, predictive analytics can be used to examine raw data and successfully diagnose equipment issues days, weeks or even months before failure. This presents a virtual goldmine – metaphorically, of course – for mining operations, with initial cost reduction and productivity gains of an estimated 10–20%, according to AVEVA. Similarly, it can help ensure that operators don’t waste time maintaining equipment that doesn’t need attention, reducing asset downtime and expenditure. Combined with a deep learning approach, predictive analytics can even forecast the remaining lifespan of an asset.

IBM has implemented this technology for the best part of the past decade, providing data analytics for mining giants like Thiess, with the goal of saving its clients up to billions of dollars each year. Back when they first introduced predictive analytics into their portfolio, all the way in 2014, Matt Denesuk, then manager of smarter planet modelling and analytics for IBM Research, claimed that the data analytics could transform the $5trn business of operating mining equipment. Beyond helping to minimise downtime and optimise maintenance schedules, predictive maintenance can help users spend less time searching for potential problems, providing early-warning indications of when an asset’s current operation deviates from the norm – even highlighting the main factors responsible for this deviation.

According to a Deloitte report, transitioning from a reactive, condition-based maintenance strategy – in other words, acting only when it becomes apparent that an asset is underperforming or in need of maintenance – to a more data-driven proactive approach can offer considerable financial savings. The company has estimated that the use of predictive maintenance in the mining industry could reduce maintenance planning time by 20–50% and overall maintenance costs by 5–10%.

Deloitte also notes that moving from a reactive, condition-based maintenance strategy to a more data-driven proactive approach can offer big savings. It has been estimated that predictive maintenance can reduce mining operations’ maintenance planning time by 20–50% and overall maintenance costs by 5–10%.

Mining operators have already implemented predictive analytics to save millions in averted asset failures, according to Fernanda Martins, process industries expert, AVEVA, in a February 2022 interview with Mining Magazine. Syncrude Canada, for example, reportedly saved $20m in annual operating cost avoidance by using this technology.

Similarly, Votorantim Cimentos, Brazil’s largest cement manufacturer, avoided $5.5m in corrective maintenance costs per site across six sites after introducing a predictive analytics solution to reduce the overall cost of maintenance, increase productivity, and enhance operational reliability. Furthermore, between 2019–21 the company saw a 10% reduction in maintenance costs and a 6% improvement in asset reliability. A 2021 GlobalData survey found that 75% of mining companies had made at least minor investments into predictive maintenance, and 48% expected to invest in the technology for the first time or invest further by 2023.

“Further investment in predictive maintenance is critical for mines looking to improve productivity and reduce expensive downtime,” said David Kurtz, director of analysis, mining and construction, at GlobalData, in the 2021 report. “The technology not only ensures continued productivity of critical operations, but saves money in parts and labour, and can even extend the life of equipment.”

The environmental algorithm

In the past, meaningful data analytics were out of reach for most mining companies for a variety of reasons, ranging from the lack of digitalisation and the resulting presence of analogue assets that could not capture data to the patchy connectivity in underground operations and a lack of a clear ROI on data-driven decision making – all of which offered little incentive for large-scale adoption. Instead, this technology was adopted and developed in other sectors, such as manufacturing, which went on to enjoy the downstream benefits.

The technological landscape has now changed, however, and as digitalisation continues to slowly seep into mine processes, so too do the benefits that predictive analytics can provide. Beyond equipment and machinery maintenance, predictive analytics also offers solutions to the ever-evolving ESG environment – particularly valuable as the effects of climate change make themselves known. Water stewardship and biodiversity are fast becoming major issues for the industry, particularly in the so-called Lithium Triangle of Argentina, Bolivia and Chile, where lithium mining relies on the evaporation of brine.

Improving a mining company’s ESG performance and accountability framework is becoming increasingly important in mining districts around the world. Accurate and up-to-date data capture and analysis are essential in order to better assess risks and opportunities in this area, and articulate these through transparent, out-based measurement and assurance. Operators that achieve this can get an edge on competitors in many ways – from accessing capital to securing a licence to operate and attracting talent. One way to ensure better, faster decisions, however, is to increase investment in data capabilities.

In 2022, many mines have implemented some form of daily and monthly reporting based on the data obtained from their operational processes. Mining company Barrick Gold, for example, was able to reduce environmental permit deviations by 45% after using predictive analytics to help ensure environmental compliance. In terms of water conservation, the mining industry now operates in an environment in which water is highly regulated, experiences unforeseen supply shocks and carries substantial social value. By 2024, water-operating expenses are estimated to increase by a 1–4% compound annual growth rate (CAGR), according to Wood Mackenzie. The mining industry has made considerable investments – an estimated $15bn in 2019 alone – to reduce water withdrawal and increase water efficiency in operations.

A central part of this is water reuse, where water is reclaimed after processing and treatment – helping to address low water availability in stressed areas and greatly reducing expenses. It’s here that predictive analytics can best be put to use, enhancing automated routines and optimising plant performances. Anglo-American, for example, has pledged to adopt techniques that will allow the company to reuse around 80% of the water throughout its mining operation, saving an estimated $15m per year.

Another key area that predictive analytics can help improve is in air pollution and mine ventilation – a key aspect of any underground mine, as it’s vital for safe operations. Deep beneath the ground, the air quality is often negatively impacted by vehicle emissions, heat and humidity, and toxic gases. In recent years, ventilation-on-demand (VOD) has been used to monitor air quality in underground operations, enabling mines to adjust ventilation according to the specific activities taking place at any particular time.

However, when paired with asset radio-frequency identification (RFID) tagging and tracking technology, VOD can do more than improve air quality and mine safety – it can help improve energy efficiency too, helping to reduce a mine’s environmental impact. This approach to underground ventilation is becoming more widespread across the industry as sensor technologies, IoT devices and the resulting data analytics capabilities have now reached a sufficient level of maturity.

As with any technology, predictive analytics will become more advanced and beneficial to the mining industry the more it is implemented – and, indeed, the global predictive analytics market is expected to grow substantially in the coming years, with projections valuing it at $28.1bn by 2026, according to MarketsandMarkets Research.

While more risk-averse mining companies might hold off on embracing this emergent technology, they will risk missing out on the benefits as it becomes more widespread. And in today’s highly competitive market, with every operator looking for an edge over their competitors, this choice could mean the difference between life and death across the industry – where a company’s long-term financial well-being is concerned, of course.

This article first appeared in World Mining Frontiers magazine.