• June 11th, 2019
  • 14 min

How Digital Technologies are Transforming the Mining Industry

As the population grows, demand for mined materials increases.  At the same time, mining these valuable materials is becoming more difficult – fluctuating commodity prices are squeezing profit margins, easily-accessible deposits are being rapidly depleted, and productivity in the sector has decreased by 28% over the last decade.  The industry is at an inflection point, and new technologies provide an opportunity to change its course.  Researchers predict that harnessing the value of digitalization can generate a potential economic impact of about USD $370 billion per year by 2025 – a figure amounting to 17% of the industry’s cost base.

An Accenture study identified technological transformation as a significant growth strategy for mining companies – 92% of mining executives plan to increase investment in technology over the next three years, and more than a quarter expect those increases to be significant.  According to a KPMG survey, the highest level of investment is occurring in data and analytics tools (53%), autonomous vehicles (30%), and robotic process automation (29%).  In this article, we will discuss a few of the digital initiatives taking place in mines around the world – from smart blast design to health and safety initiatives – then explain how mines can overcome common barriers to unlocking the full potential of these technologies.

Blast Better to Drive Efficiency

Even though the industry has been talking about mine-to-mill optimization for the last 30 years, we are nowhere near realizing its full potential.  Comminution accounts for 53% of the energy consumed on site and, as ore grades continue to decline while the cost of processing increases, we can only expect this figure to increase.  We know that particle size is the key metric to analyze, and that optimizing particle size at each stage can increase throughput by up to 30%.  But the reality is that traditional sieve analysis and early image-based processing techniques are too labour-intensive for mines to implement.  Technology can change that.

Mine-to-mill optimization requires continual monitoring at each stage in the comminution process.   Motion Metrics offers particle size analysis solutions for shovels, on conveyor belts, and in a portable format that don’t interrupt production.

Blast optimization is the low-hanging fruit of the mine-to-mill optimization process – milling is about ten times more expensive than blasting.  You will not get anywhere by blindly increasing blast intensity but monitoring particle size to tailor your blast parameters can dramatically improve performance.  At Motion Metrics, we offer two ways to analyze your blast – one is an in-shovel particle size analysis solution and the other is a handheld, point-and-shoot device.

Both solutions use deep learning algorithms to provide fast and accurate particle size distributions without the need for a scaling object, but they have different use cases.  An in-shovel solution will provide a higher sample rate for regular coverage, whereas a handheld solution provides flexibility.  For example, a mine will need a handheld solution to evaluate the particle size distribution of a salvage pile or a stockpile being loaded by small excavators where an in-shovel system has not been installed.  Smaller mines that blast less frequently may not need both solutions, but large operations often benefit from comprehensive coverage.  For these cases, we recommend supplementing an in-shovel solution with a portable analysis tool.

Further down the comminution chain, we offer a particle size analysis solution for conveyor belts that uses stereo imaging and deep learning to provide 24/7 monitoring.  This solution can also detect when the conveyor belt is empty, alerting mine personnel to blocked screens or chutes.  Pairing conveyor belt analysis with in-shovel and portable solutions provides the in-depth insight into blasting, crushing, and grinding that mines need to continually optimize.   Collected data could even be used to automate workflows – for example, adjusting the crusher gap automatically.

Pairing conveyor belt analysis with in-shovel and portable solutions provides the in-depth insight into blasting, crushing, and grinding that mines need to continually optimize.  Our solution uses artificial intelligence and 3D imaging to analyze particle size 24/7.

Any of these solutions can provide an impressive return on investment.  For example, a Teck coal mine installed our shovel-based particle size analysis solution to study the relationship between powder factor and excavation efficiency over a two-year period.  The mine, which blasts 80 to 100 million cubic metres of waste rock each year, iteratively reduced their powder factor while measuring the impact on particle size distribution and shovel loading times.  To measure the impact on particle size distribution, the mine sieved the blasted material and determined the screen size through which 80% of the particles would pass (P-80).  By the end of the study period, the mine had reduced their powder factor by 15% and increased P-80 by more than 20% without impacting the efficiency of excavation.  By continually analyzing particle size, the mine optimized their blast performance to save 8.6 million kilograms of explosives.

Particle size optimization is a natural starting point for technology-driven mine-to-mill strategies, but there are many more opportunities to improve along the comminution chain.  Material transport is one such opportunity that mines around the world are beginning to exploit.

Make Haulage Fleets More Efficient

With all the press that Uber, Google, and Tesla have received for their progress towards autonomous cars, you’d be forgiven for missing what’s going on in the mining industry – Rio Tinto has been using autonomous haul trucks at its Mine of the Future since 2008!

Manufactured by Komatsu, this fleet of 71 haul trucks uses a combination of GPS, radar, and LIDAR to navigate completely independently around Rio Tinto’s Pilbara site in Western Australia.  Because the trucks can operate nearly 24/7, stopping only for maintenance and to refuel, Rio Tinto has already seen cost savings. The autonomous fleet outperforms the manned fleet by an average of 14% and has reduced operating costs by 13%.  The trucks currently move about 20% of the operation’s materials, but Rio Tinto plans to double its fleet by the end of this year.

Rio Tinto has been using autonomous haul trucks built by Komatsu at its Mine of the Future since 2008.

BHP has also deployed autonomous haul trucks at its Jimblebar iron ore mine in Western Australia.  This change has reduced costs by about 20%.  Although the bulk of media coverage has focused on personal transportation, the mining industry is a more suitable candidate for early deployments – there are fewer humans present, workflows take place in a controlled environment on privately-owned land, and mining operations involve repetitive tasks with minimal variability.

Although autonomous vehicles have had early success on site, mines do not need to fully automate their haulage systems to see big benefits from technology.  A Congolese copper mine hired a team of data scientists to quantify the impact of operator behavior on fuel consumption and saved big.  Using low-cost spatial trackers and drones to track truck dispatching and vehicle movements, the scientists used specialized statistical software and deep neural networks to determine how operators were driving their haul trucks.  With this information, the scientists were able to limit peak speeds, reduce short stops and restarts, and avoid abrupt braking and strong accelerations.  The result?  Fuel consumption dropped by 7% in just eight weeks.

Another way that mines can use technology to squeeze more value out of existing equipment and infrastructure is to invest in predictive maintenance.  These solutions can help mines determine the condition of connected assets and predict when maintenance should be performed.

Improve Maintenance Efficiency with Better Planning

Predictive maintenance solutions that address problems before they occur are one of the biggest benefits that digital technologies can offer the mining industry.  Maintenance in mining often occurs on a routine or time-based schedule, rather than as needed – leading to a lot of wasted time and money.

Performing maintenance only when warranted, rather than on a fixed schedule, can help mines better allocate resources and save money.  One rail operator, for example, enlisted a team of data scientists to conduct a five-week machine learning proof of concept.  The project culminated in a customized optimization tool that would help workers identify ballast areas in need of remediation.  Using the optimization tool, the operator eliminated unnecessary maintenance to reduce ballast cleaning costs by up to 13%.

Brazilian mining giant Vale uses artificial intelligence to predict rail fractures on the Carajas and Vitoria-Minas railroads so that these incidents can be prevented.  The initiative has helped reduce the occurrence of fractures by up to 85%, which can save Vale $7 million per year.  The company expects to save about $26 million in 2018 from this project alone.

Even mine railways can benefit from digitization.  Today, mines are using artificial intelligence to predict rail fractures and combat ballast fouling.

At Motion Metrics, we use machine learning to help mines implement predictive maintenance for shovels.  Our smart solutions provide tooth wear monitoring for all mining shovels and excavators so that tooth change-outs can be predicted and scheduled without the need for manual measurements.

Excessively worn shovel teeth pose a few problems to mines.  Firstly, worn teeth are inefficient – they require increased cutting forces to penetrate material and result in lower overall productivity and increased energy consumption while causing premature wear to other sacrificial components like adapters.  Operating with excessively worn teeth also increases the risk of breakage – and missing teeth can travel downstream to jam crushers.

Traditional methods of tooth wear monitoring are time-consuming and require expensive downtime.  Our machine learning-based method analyzes tooth length continually without interrupting current workflows.  Reports can be viewed from our centralized data analysis platform.

Finally, unplanned tooth change-outs are far more expensive than scheduled ones.   The costs of tooth replacement are both direct and indirect.  Direct costs include the teeth and labour; these are generally fixed and relatively small.  Indirect costs refer to lost production caused by shovel downtime and depend on the given operation’s production rate and net profit.  When tooth change-outs are scheduled, other maintenance tasks can be performed simultaneously to reduce the indirect costs of maintenance.  Therefore, the incremental cost of an unplanned tooth change-out is significantly greater than a scheduled one.

According to an earlier study, the direct and indirect costs due to an unplanned tooth set change-out are about USD $3000 and USD $38,000 respectively; together, the extra costs and lost opportunity amount to USD $41,000 per incident.  Although unplanned change-outs are sometimes unavoidable, these incidents can be significantly reduced with a tooth wear monitoring solution.  Since shovel downtime is the most expensive part of tooth change-outs, a small investment in replacement interval optimization can save mines a lot of money.

In addition to reducing costs, predictive maintenance technologies can also advance the most important goal of all – getting mine workers home safely.  Keeping mine equipment and infrastructure in good shape reduces the risk of fatalities, injuries, and environmental hazards like dust and noise.  In the next section, we’ll explore some of the ways that mines are using technology to improve worker safety.

Get Serious About Safety

When it comes to digital technologies, the financial opportunity is clear – experts predict that using smart sensors on site could translate to savings of USD $34 billion by improving equipment utilization, reducing equipment failures and downtime, and facilitating predictive maintenance.  But lives are always more important than dollars, and the opportunities to improve mine health and safety with technology are massive – the World Economic Forum predicts that digitization could prevent 44,000 injuries and save 1000 lives over the next decade.

Automation is the most obvious way to keep mine workers out of harm’s way.  In the long term, researchers estimate that automating operations can reduce the number of people working in dangerous areas by more than 50 percent.  In the short term, there are many applications for modern technology on-site that don’t interrupt existing workflows or change jobs and that can be easily implemented today.  For example, BHP trialed a program to measure and act on truck driver fatigue at its Escondida copper mine in Chile.  BHP outfitted more than 150 trucks and drivers with smart caps that analyzed driver brain waves to boost productivity and increase safety.

Similarly, artificial intelligence can be used proactively to detect hazards like toxic gases, dusts, and radiation on site.  Imagine a series of preinstalled monitoring stations that inspect the worksite ahead of mine personnel by using robots, sensors, and artificial intelligence.  These systems could trigger alarms, warn workers, and close the affected area to prevent further exposure to hazards.

We are also working on preventative safety solutions at Motion Metrics.  While it may not be possible to avoid incidents entirely, efforts should be taken to mitigate danger as early in the workflow as possible.  This is the mentality behind our flagship missing tooth detection solutions for shovels and loaders, which have been deployed at mine sites around the world for more than 15 years.

Our missing tooth detection solutions use artificial intelligence and a rugged camera to detect missing teeth before they can be transported downstream with the excavated ore.

Shovel and loader teeth are prone to breaking off during operation.  These missing teeth can go undetected by equipment operators and be transported downstream in the excavated ore where they can obstruct crushing equipment.  An obstructed crusher always presents a serious safety issue for any mine due to the tremendous amount of stored energy.  American researchers at the National Institute for Occupational Safety and Health found that incidents involving crushers are the second most common cause of fatalities caused by stationary machinery at mines in the United States.

To free an obstructing tooth, all other material in the crusher must first be removed with an excavator.  Then, a welder uses a thermic lance or torch to heat and cut the obstruction to loosen it.  Because heat causes the metal to expand, the lancing process generates additional pressure that can cause the rapid, uncontrolled release of the tooth.

Not every broken tooth will obstruct a crusher, and not every crusher obstruction will result in a fatality or an injury.  However, clearing a crusher obstruction requires significant downtime that results in massive production loss – avoiding even a single crusher obstruction pays for the cost of a missing tooth detection solution.  For example, a Chilean copper mine installed our missing tooth detection solution for loaders and, over a 10-month period, detected 12 missing loader teeth.  The mine experienced zero crusher downtime and, according to conservative estimates, avoided a production loss of about 144 hours.  Over the long term, these systems are an inexpensive way to reduce the risk your team is exposed to.  Viewed through this lens, a missing tooth detection solution is an insurance policy well worth having.

Break Down Barriers

Now we get to the big question – if emerging technologies are so game-changing, then why are they not being implemented at scale?  Motion Metrics has been at the intersection of mining and technology for years and we have identified some of the common pitfalls driving this disconnect:

  1. Ease of use. New technologies can be expensive and time-consuming to implement because employees must be trained to use the new systems.  Although all new technologies require some degree of organizational learning, those that place high demands on workers will be underutilized.  We’ve learned to build solutions that are intuitive and require minimal training.
  2. Unclear implementation. Agreement on a digital vision is not enough – there needs to be a clear plan to move from the current state to the end goal.  Decision makers should take the time to draft a detailed implementation pathway that identifies stakeholders and includes detailed timelines, deadlines, and key performance indicators for success.
  3. Perception of costs. There is a common misconception that mining companies are technology averse.  They are not.  Risk turns mining companies off – in an industry where efficiency is king and investments that under-deliver can make the difference between a profitable operation and an unprofitable one, mines can only invest in solutions with a clear return on investment. Choose initiatives wisely and ask your technology provider for numbers.
  4. Lack of communication. The owners of digital transformation are often unclear at mines.  Why?  Corporate innovation departments purchase new technologies, but on-site teams use them.  To drive value through technology, decision-makers need to create accountability, break down silos, and open channels for communication between teams.

When it comes down to it, the barriers to widespread implementation are not technical – they are about people.  Solution providers need to design products that are as intuitive as they are powerful, and miners need to treat innovation initiatives with the same level of attention and detail that they would any other aspect of their operation.   This is new ground for both mining and technology companies, and success will depend on strong partnerships and effective collaboration.

You might also be interested in
See all posts