Automation, AI and machine learning in mining: What is the reality?

One would be forgiven for thinking that automation and robotics would have been driven by mining, due to our remote mine sites, the hazardous nature of work and the high costs of labour and transport.

However, like many leaps forward in technology, it was most likely driven by war and conflict, due to a severe shortage of labour and the pressing need to build more tanks, ships and planes. Hence, manufacturing trumps mining as the birthplace of industrial automation.

Back in 1962, the first automation process was installed at General Motors, with the convergence of two main technologies: the Electric Servo Motor and Numerically Controlled Machines (punch cards).

In today’s mining operations, automation is possible due to the convergence of quite a number of technologies, including the advancement of GPS technologies, machine learning, wireless communication and the ability to store and transfer huge quantities of data. While automation is not necessarily completely new, it is more likely that the collaboration of technologies has opened up new opportunities for its application. After all, self-driving tractors have been used to plough fields for centuries; we call them oxen.

The primary concern with automation is, and always has been, what are the human consequences? Commentary on this ranges wildly from utopian to catastrophic scenarios, and we would be advised to focus specifically on how technology and automation will be used in our specific industry or operation.

Let’s begin with the ‘why’ of automation. The most commonly used reasons include:

  • Safety: Removing humans from hazardous environments or activities and eliminating ‘human error’
  • Efficiency: Completing repeatable tasks in an accurate and timely manner for an indefinite period of time
  • Reliability: Availability and consistency of the desired outcome.

In the majority of cases we are talking about automated tasks and the net benefit to humans in simple terms has been new opportunities for employment in technological industries, whilst eliminating mundane activities. Think of the dishwasher, GPS, data roaming, or temperature control.

As we move from task to decision-oriented automation, such as artificial intelligence and machine learning, there may be the perception that this impinges on our unique realm, and brings into play the uniquely human concept of trust. Trust and value in human and machine interactions should be more thoroughly investigated by managers and recruiters.

Regardless, trust suggests that reliability and automation may, over time, shift the primary focus of mining operations from equipment reliability to system reliability.

As more and more sensors are placed into equipment across a minerals processing plant, and more data is captured and used for critical decision making, then the availability of the network is paramount. The ability to store data and the integrity of information along the entire supply chain should be the first foundation piece before automation is embraced.

In terms of priorities, many operations implement a layered approach. The ability to produce product is fundamental and therefore mining production, maintenance planning, and reliability strategies are largely similar.

Condition monitoring, data mining and machine learning will move the needle to assist with predictive maintenance, auto fault diagnosis and automated parts ordering. Similarly, process optimisation can be automatically streamlined based on the mine data received.

Network outages or breaches will not necessarily be devastating; operations can easily continue but as the processes becomes more interconnected and complex, the next layer will focus on the network and the final layer on data integrity and security.

Full scale integrated automation should drive collaboration in a way we have not really seen in our industry.

Equipment suppliers, services providers and transport partners will need to ensure they are delivering open architecture condition monitoring, sensing, data logging, and communication that suits the mining and minerals processing operation in order to deliver a homogenous system.

A number of shipping and rail transport companies are well ahead of the game in terms of their ability to monitor and predict current and future circumstances that may feed into the mining operation. Collaborative and trusting relationships at a business level, and at a network level, will be the key to maximising the value of automation in mining here in Australia.

In terms of human impact, will automation and artificial intelligence eliminate or reduce the need for critical thinking or human contribution?

Not at all, there are many references looking at the evolution of human technology relationships in the mining industry.

We certainly may see a reduction of machine operators located at the mine site over time but, we will see an increase in demand for control systems integrators, network engineers, data engineers, programmers and data analysts.

In fact, I believe that the full adoption of new technologies, automation and robotics will align with the skills and experiences of the new generation of graduates and future employees, making mining a more attractive career choice.

To keep up to date with Australian Mining, subscribe to our free email newsletters delivered straight to your inbox. Click here.