Mining moves up the digital maturity curve

They are exploring and investing in analytics and artificial intelligence (AI) in a bid to leverage the data they generate, and use it to improve planning and decision-making across the mining value chain. In the increasingly digitised environment, Deloitte defines three horizons of AI in its 2019 Tracking the Trends report – assisted intelligence, augmented intelligence, and autonomous intelligence. The maturity along this digital path varies, often considerably, between mining companies and in comparison with other industries. In a 2018 global Deloitte survey, energy and resources organisations rated themselves on average between three and four on the maturity scale, compared to an average of between five and six in other industries. Many companies are already working in the realm of assisted intelligence, which includes robotic process automation, typically targeted at repetitive back-office processes. It has been the global mining companies, such as those with Pilbara iron ore operations, that have been more likely to venture into augmented intelligence or autonomous intelligence, however. Augmented intelligence is when machine learning augments human decisions, while autonomous intelligence is when AI decides and executes autonomously. The urgency for other organisations to make moves into AI and analytics has risen as market conditions have improved off the back of rising commodity prices. There has also been increased visibility of the potential of new digital capabilities being applied in other industries. Deloitte Consulting partner Paul Klein identifies two key themes that have emerged as this digital frontier has started to unfold. The first focuses on where mining companies should start and which investments will deliver the best returns. “Is it targeting operations first or the back office processes? Is it targeting particular parts of the value chain where there are current problems? There is discussion and debate about where the biggest return is, and this varies between organisations.” Klein tells Australian Mining. The second theme, and perhaps the one the industry grapples with most, centres on the impact increasing automation and a reliance on machine intelligence will have on workforces. Klein says the simplistic view is that automation and other technologies will replace jobs, but this is often not the case. “Yes it does replace some jobs and for good reasons because we are taking people out of dangerous situations, or we are taking people away from repetitive back office tasks that are better done by machine intelligence,” Klein says. “In the majority of cases, machine intelligence is augmenting human intelligence. “People and machines are working together to solve problems, and analytic algorithms are suggesting, prompting and improving human decisions. “They are not replacing the human altogether, which is something that is often misunderstood in this space.” Klein expects future examples of horizon two and three will have an important human involvement that will instead require companies to upskill workers. Deloitte works with companies making this shift, including upskilling and sourcing workers to manage and analyse information from across the value chain. “Using analytic algorithms to improve and augment human decisions requires skills around data management and data science,” Klein says. “Skills around the analysis, interpretation and visualisation of information are relatively scarce and in high demand. “A number of mining companies are making investments in hiring people with those skills and then creating a mix of those people alongside others with more traditional mining backgrounds to make the most of that capability.”

That’s a culture shift that I think organisations are only just getting into – the willingness to try new things and accept they are not all going to work

Deloitte suggests companies take a three-stage approach to implementing analytics and AI solutions – think big, start small, scale fast. To ‘think big’ miners should “establish a clear vision and strategy, driven by desired business outcomes.” This point relates to the earlier point about deciding where to focus investment for the greatest return. With a vision and strategy in place, miners are recommended to “design and deliver in rapid agile sprints, starting with a minimum viable product or solution.” The aim is to quickly get to the point where an initial version of a technology solution can prove it delivers value, rather than taking months to design and build a more complete solution. Finally, Deloitte suggests companies “rapidly scale up successful projects that have demonstrated value (and kill-off those that don’t), and focus on embedding the change operationally.” Klein says to succeed with this approach it will require a change of culture at some companies.  

“The traditional mindset at mining companies around the risk of new investments and how new capital projects are approached often drives the thinking that makes it hard to accept that a project might fail after an initial investment,” Klein says.

“There is often a reluctance to proceed with a project unless it has been through multiple feasibility steps. There is a bit of natural resistance in most organisations to trialling something and being willing to accept that it might fail. “That’s a culture shift that I think organisations are only just getting into – the willingness to try new things and accept they are not all going to work, and then to focus further investment on rapidly scaling up those that have proven their value.” Deloitte’s experience with successful projects that move companies along the maturity curve is that they deliver a range of outcomes across organisational structures and operational portfolios – improving safety, increasing productivity, reducing cost, or enhancing the daily experience for employees who have increased expectations due to the capabilities of consumer technologies. Klein says the use of AI and analytics in mining to date has typically been limited to targeted point solutions that deliver value in one part of the chain, or at one operation. For example, using machine learning to optimise blasting operations, using prescriptive modelling to optimise process plants, using real-time monitoring for maintenance planning, or using advanced analytics to identify patterns in safety data and gain insights (source: Tracking the Trends).

“Increasing analytics maturity requires greater integration of data from multiple sources and delivery of end-to-end planning and decision-making solutions that span multiple processes and operations, and deliver more significant benefits,” Klein concludes.