Tag Archives: machine learning

BHP, SensOre progress artificial intelligence-backed exploration agreement to ‘Phase 3’

SensOre says it is to advance its Joint Targeting Agreement (JTA) with BHP to “Phase 3” after receiving approval from the major miner.

Under the JTA, SensOre was required to meet certain hurdle rates and technical thresholds through deployment of its Discriminant Predictive Targeting® (DPT®) technology and related auxiliary systems. SensOre says it has met or exceeded the requirements set for Phases 1 and 2.

Richard Taylor, CEO of SensOre, said: “The SensOre team has been excited by the performance of its systems in targeting new commodity and deposit types. The relationship with BHP and its support for innovation in exploration has been incredibly valuable. The results derive from the truly joint nature of the project and shared view that better use of geoscience data will lead to improvements in discovery rates. We are really thrilled with the results.”

SensOre and BHP reached agreement on a letter of intent in May 2020, confirmed via execution of the JTA on September 18, 2020. The JTA envisages a phased process, training the DPT technology on commodity-specific deposit types and applying the knowledge gained to a predetermined search space. SensOre stands to benefit from fees for the targeting exercise and potential success-based payments on certain discoveries arising from the technology, it said.

SensOre aims to become the top performing minerals targeting company in the world through the deployment of artificial intelligence and machine-learning technologies, specifically its DPT workflow. SensOre collects all available geological information in a terrane and places it in a multi-dimensional hypercube or Data Cube, with its big data approach allowing DPT predictive analytics to accurately predict known endowment and generate targets for further discovery, it says.

Searcher brings GeoClerk geo-imagery search engine to mining sector

Searcher is looking to branch out of the oil & gas sector with a mining edition of its advanced geo-imagery search engine GeoClerk.

GeoClerk intelligently extracts imagery and surrounding data from all types of unstructured documents, and then uses machine learning to classify them into geologically relevant categories.

The intuitive, web-based interface enables simple and complex querying to provide images in categories such as maps, core photos, imagery, cross-sections and tables, which are all linked back to the original source documents, the company says.

GeoClerk works with government databases, mineral company websites, internal document libraries and institutional subscriptions to identify, define and present images in geologically relevant categories to easily extract information from historic and uncategorised documents, Searcher says. Over 1.2 million relevant images are already available in GeoClerk.

Helen Anderson, Vice President Minerals at Searcher, said: “Expanding into the mining industry, after 15 years of oil and gas focus, is an exciting new venture for Searcher.

“Industry feedback has been overwhelmingly positive since the launch of GeoClerk in the oil and gas industry. There was a clear demand for geoscientists from the mineral industry to utilise the geo-imagery search engine.

“Companies in both industries are looking to save time and resources, as well as making it easier to access their huge databases of old reports in storage units and modern computerised data, all through the visual aspect of images.”

Maptek looks back on 40 years of mining software advances

Maptek is looking back on its roots, 40 years after geologist Bob Johnson laid foundations for the company to become a leading provider of innovative software, hardware and services for the mining industry.

In the mid-1970s, Johnson opened a small bureau service above a row of shops in suburban Sydney, New South Wales, Australia, to computerise coal seam drafting. That venture was the precursor to Maptek, which today develops, sells and supports innovative mining solutions to more than 20,000 users worldwide.

In 1981, Johnson then formed a company to allow customers to do their own computer work. That became Maptek, which today employs 350 staff in 18 offices to support a customer base including the world’s biggest mines across more than 90 countries.

“The transformation from startup to global technology developer did not happen overnight,” Maptek founder Johnson acknowledges, as he reflects on what defines Maptek today. “Innovation results from many small increments – it rarely happens from an epiphany.

“We started off by computerising the plotting of boreholes and mapping of coal deposits, which, until then, was a very tedious manual process. People were asking if it worked for all commodities, not just coal, and I realised we needed to put the software in the hands of the users. This was how Maptek came about.”

Johnson states that Maptek sets and continuously strivers to hit a high standard.

“Early computing in the 1980s was the breeding ground for automating manual tasks and it was a challenge to convince some people to replace existing practices,” he said. “Tradition dies hard!

“Maptek integrated multiple steps in the computerisation of mining applications. In this way we were able to own the workflow and it’s probably key to why our first customer, BHP Coal, remains a customer today.”

He added: “Do something different and stay in front is a guiding principle that remains a key business value for Maptek.”

Fast forward to 2021 where CEO Eduardo Coloma is embracing the vision, with a long-term technology development roadmap to deliver state-of-the-art solutions and exceptional customer experience, the company says.

“Maptek intends to stay ahead by continuing to be a disruptive influence and affect change for the betterment of the mining industry,” Coloma says.

The new Mining 4.0 paradigm has five characteristics, according to Coloma.

“Vast amounts of data; delivering that data to the right people at the right time; efficient data storage and universal access to it; using technology for computationally-intensive tasks; and data-driven decision making…all need to be balanced,” he said. “Add to that the challenges that the pandemic unleashed!”

He added: “With challenge comes opportunity. Miners are continually on the lookout for smarter processes.

“Maptek was conceived 40 years ago at the start of the digital revolution. Customers today have an ever-growing appetite for technologies to enable digitalisation and automation. They are not afraid of new technology and look to us to lead them.

“It’s not just technology that is fast-evolving, the people and organisations who consume it must also be open to adopting new ways of working. Digitalisation has provided the conduit for data to be universally accessible and dynamically updatable.

“We want to make sure our customers get the most of their data, sharing it across the organisation in such a way that everyone benefits. Data is being democratised!”

A data-driven culture embraces systems which are robust, repeatable and user-independent, according to Coloma.

“Crucially these systems meet the needs of a mobile, shift-based and geographically dispersed workforce,” he said.

“We build technology solutions that allow our customers to turn their data into knowledge and use that knowledge to support business improvement. We provide an automated decision support ecosystem…they provide their individual experience and intuition to make that knowledge relevant to their business.

“Already we are exploiting machine learning and digital twinning to connect the planning cycle to production performance data for comparing performance against plans.”

With fewer barriers to extending technology within mines, companies are looking at the entire value chain to make improvements. Maptek can help connect processes, functions and data to enable more accurate, predictable and profitable operation of mines, it says.

In closing, Coloma explains why Maptek is well placed to help mining companies use their data as a bridge to continuous improvement.

“Our unique culture, instilled by our founder Bob Johnson, gives staff a great amount of freedom to be innovative,” he said. “It fosters imagination everywhere and is the key to continued success.

“We give our customers the freedom to dream and ask for solutions to their real world problems.

“Our enduring relationships with customers are hugely important in our ability to solve these challenges. Bob mentioned our first customer, who remains a customer today. But accepting that change is inevitable is a reminder to us not to rest on tradition.”

MICROMINE makes a software splash at Diggers & Dealers

With Western Australia one of MICROMINE’s key markets for its Micromine and Pitram products, it is hardly surprising the software leader chose this week’s Diggers & Dealers Mining Forum in Kalgoorlie to reveal a host of new updates for the 3D modelling and mine design/mine production and fleet tracking solutions.

Across the company’s product suite, MICROMINE has been readily engaging with customers throughout the world, with users providing feedback to form its product roadmaps.

One of the results of this consultation process is a move to a six-monthly release cycle to enable its software to grow and adapt with clients’ operations.

Another is providing networking options to expand usage of its software across a wider number of users – the free Micromine Effects reader enabling anyone to view, share and interrogate Micromine output files without needing access to a full software licence.

“We’ve also introduced subscription offerings which our customers have quickly adopted because they provide a flexible and scalable option for large teams to access more functionality across our product suite, with less upfront cost compared to the traditional perpetual model,” Adam Brew, MICROMINE Australia Manager, told IM.

Shifting any capex item to the opex column is bound to go down well with the mining community, as MICROMINE has shown.

Having occurred in August 2019, the move led to almost nine months straight of subscription-only sales, according to Brew. “It surpassed our expectations,” he said. “The ability to have a subscription model allowed us to then launch the Free April campaign.”

The “Free April” campaign – which saw MICROMINE offer miners complimentary access to its general mining Micromine package during April as COVID-19 started to bite – led to around 4,000 new people interacting with the software, according to Brew.

MICROMINE has been busy on updates during the pandemic, but it has also delivered its first fully remote implementation of Pitram at a mine operation in Greece, leveraging the experience from its global Pitram support desk to fully deploy a Pitram FMS and Material Management solution.

This Greek project is well advanced with Pitram playing a crucial role in a major refurbishment and expansion of existing operations. The solution at the mine is aimed at helping improve development and production mining cycles; accurately track materials from source to processing; provide Online Analytical Processing reporting and analysis; enhance reactions to, and minimise the impact of, unplanned events; and increase equipment availability and utilisation.

Yet, those attending the MICROMINE booth at Diggers & Dealers this week will have even more to talk about.

Something new

“Micromine 2021 is scheduled for release later this year and attendees of Diggers and Dealers will be the first to get a pre-release reveal of our flagship software offering,” Brew said.

Australia, in particular, has bucked global trends in terms of exploration expenditure, and the Micromine value proposition has been central to the company capitalising on this resurgence in exploration activities, according to Brew.

It is no wonder then that the company has put significant efforts into updating its flagship product.

“The first thing clients will notice is a completely redesigned user interface that provides easier access to the critical functions of the software, transforming the whole user experience with responsive design and efficient workflows,” Brew said.

Delivering this transformation has been a focal point for the business for more than a year, according to Brew, with developers reviewing customer requests most commonly received from the support team, analysing how users work with the array of Micromine functionality, and modelling interface scenarios to optimise the presentation of key functions within the software.

“By providing easier access to these functions and a smart interface that responds contextually, Micromine 2021 anticipates and supports workflows in a genuinely intuitive way,” Brew said.

The Micromine update has more than a new look.

It also includes new tools for importing and working with as-drilled drill-hole data, Brew explains.

These provide faster and more intuitive control over underground ring drill and blast design – also a focus of the earlier Micromine 2020.5 update – enabling designs to quickly adapt to changes in the field, identifying drilling inefficiencies and improving design protocols.

“We are also introducing intuitive tools that mirror the terrain of a blast face and speed up the process of creating blast-hole patterns within the bounds of the dig block,” Brew said. “Users will be able to accommodate polygons/blast masters of varying shapes, reducing the need for manual adjustment.”

The new grade control capabilities in Micromine 2021 provide dynamic updating of grade control reports to enable faster design preparation and reserve evaluation, according to Brew. This can allow miners to explore variations in dig block configuration and evaluate the ramifications of design changes on the grade – a function bound to appeal to opex-focused companies mining complex orebodies.

An integrated scheduler, meanwhile, enables planners to build and visualise an optimised schedule through configurable templates, scripting capabilities and scenarios built from real-world constraints, Brew said.

While the new and intuitive interface is likely to capture the immediate attention of users, MICROMINE has evidently not scrimped on updated and upgraded features.

Getting to the core

With the release of Pitram 4.17 earlier this year, there were improvements to the Materials Movement and Shift Planner modules, but Pitram 5, to be released later this year, goes above and beyond that.

“Stockpile management is now part of your end-to-end process and not managed as isolated assets within Pitram,” Brew says of Pitram 5. Geologists can work with data up- and down-stream to manage and react to material mismatches. Such data validation and accuracy is key to the value proposition Pitram drives in MICROMINE’s global implementations, according to Brew.

“Pitram is at the core of any mining operations ecosystem,” he said. “Our ability to accurately track Last Source, Destination Moved, Quantity and Grade as well as set individual depletion models across the various stockpiles across the mine, makes it a more flexible offering while maintaining data integrity.”

This near real-time tracking ability has previously failed on occasion from connectivity issues.

Not anymore.

“Pitram 5 is a huge leap forward in how we deploy our solution from a connectivity point of view,” Brew said. “Many of the mines we work with have limited or varying degrees of underground Wi-Fi and communications available. Our Peer to Peer solution bridges the gap where communication back to the server is not available at the face, for example.”

The Peer to Peer software can be installed on light vehicles which move around the mine encountering heavy equipment and collecting data in areas of no network coverage before moving back to a Wi-Fi-enabled area to sync the data back to the main server and into the control room. This allows miners developing new areas of their operation to keep up the communications flow without the need to immediately install or expand a communication network.

Such a solution has been successfully deployed at several sites globally, with Independence Group’s Nova nickel operation, in Western Australia, being the company’s reference site.

“Additionally, we have driven more R&D in how we can better leverage our Pitram Restful Integration Service (PRIS) to communicate shift planning data back to the shift bosses and mine managers in near real time,” Brew said.

The free Pitram Connect application, downloadable from the Apple or Google Play store, will show users real-time shift data as well as give them the ability to make updates to the shift, such as equipment or location allocations.

“Our ability to deliver on short interval control is a common requirement we are measured against and providing this planner to key users underground unlocks considerable value for an operation,” Brew said.

Pitram 5’s machine-learning update in the 2021 release leverages the company’s learnings from earlier deployments at some Central Asia mines.

“Utilising the processes of computer vision and deep machine learning, on-board cameras are placed on loaders to track variables such as loading time, hauling time, dumping time and travelling empty time,” he said. “The video feed is processed on the Pitram vehicle computer edge device, with the extracted information then transferred to Pitram servers for processing.”

Reflecting on the product updates and more than six months of pandemic-affected upheaval, Brew concluded: “Our business is extremely fortunate to have powered on through the COVID-19 pandemic, and we’ve worked hard to maintain our renowned ability to work, support and deploy our solutions remotely.

Diggers & Dealers is the pre-eminent event for the Australian region of our business, with representation from all our customers, so it represents a fantastic opportunity to show how we continue to drive value to our existing customer base as well as connect with new customers.”

Aspen Tech and Wood to offer clients predictive, prescriptive maintenance solutions

Aspen Technology and Wood have announced a new partnership that will offer Wood’s clients Aspen Mtell® asset performance management (APM) technology for predictive and prescriptive maintenance.

The partnership will enable global enterprises to improve the performance of their manufacturing and facility assets through a maintenance solution built upon industrial artificial intelligence (AI) and machine learning, the companies said.

Aspen Mtell analyses historical and real-time operational and maintenance data to discover the precise failure signatures that precede asset degradation and breakdowns, predict future failures, and prescribe detailed actions to mitigate problems, they explained.

Wood has decades of experience providing solution-independent asset performance consulting, as well as integrating and deploying specialty engineering services and real-time performance monitoring systems, some of which has been mining-related.

“The combination of this deep domain expertise of asset and operator challenges, with AspenTech’s extensive knowledge of the process manufacturing industry and proven AI-driven predictive and prescriptive maintenance solutions, provides a unique customised asset performance management solution for operators’ needs,” the two companies said.

Prabu Parthasarathy, Vice President of Intelligent Operations at Wood, said: “Wood has an extensive understanding of the performance optimisation needs of our clients and realised a unique opportunity to provide a solution to help enhance asset productivity and identify potential issues well ahead of time.”

Darren Martin, CTO at Wood, added: “We are excited to bring AspenTech into our strategic partnership ecosystem to unlock innovative technology solutions to solve our clients’ challenges. Aspen Mtell is part of our connected operations and maintenance programs that will allow our clients to detect patterns in operating data, allowing them to take prescriptive action and avoid unplanned downtime. Together, our vision is to drive value through digital twins across the full asset lifecycle, working to optimise asset performance, monitoring, and control across any environment.”

Greg Mason, Senior Vice President and General Manager of APM, Aspen Technology, said the value of predictive and prescriptive maintenance represents much more than simply predicting failures on large rotating assets.

“Companies that are truly focused on eliminating safety and environmental incidents tied to machine failure, in addition to avoiding production losses, understand the need to have a comprehensive predictive maintenance culture throughout the entire plant,” he said. “This requires an analytics technology that is scalable, resources needed to deploy to scale, and the expertise to lead change management. I’m pleased to say that the partnership of AspenTech and Wood around the Aspen Mtell solution provide these three unique capabilities needed to bring contextualised AI for the process industries to scale.”

DataCloud bridging the mining industry’s data divide

DataCloud is looking to collect and merge the mining industry’s datasets through a cleaning, processing, integration, and predictive analytics platform that can help different stages of an operation prepare and plan for the ore and waste heading their way.

While the coarse ore stockpile may be the section of the flowsheet currently in DataCloud’s crosshairs – thanks to a well-attended webinar a few months back – any part of the mining process that is “between departments” could benefit from the MinePortal solution, according to Steven Putt, Director of Software Solutions for the company.

“The value case is inherent anywhere between departments – ie the stockpile is after crushing, but before the mill,” he told IM.

“The reason that stockpile is there – it tends to only be half a day or a day’s material – is it is a buffer for the mill,” Putt said. “Within this pile, one truck might have been hauling very hard material that the mill is exclusively treating for a week or so. Then, in accordance with the mine plan, this can switch to another truck and a new area of the mine, meaning the mill is going to have to adapt to a completely different material.”

The distinction between material in the coarse ore stockpile is often not this apparent; it tends to represent the mine site’s ‘melting pot’, taking in material from all over the operation.

Yet, to operate effectively, the mill needs to know the origins of the material coming its way ahead of time. The mill would then, ideally, be re-configured to treat the material.

“The mill operator would need to change the speeds of operation, the water balance, potentially the grinding media, etc,” Putt said. “Operators would typically prefer not to make those changes though, having the mill running at some ‘optimal’ speed based on the idea that the material is relatively consistent.”

The reality of the situation is different, as DataCloud and its MinePortal platform have been proving.

“The last client we worked with could end up saving around $20 million a year by carrying out our recommended processes as part of a wider mine to mill tracking solution,” Putt said of a copper-gold operation the company worked at. “Basically a specific rock type (skarn) was being fed into the mix too often and the mill was not prepared to handle this in the blend.”

This client turned out to be spending more money than necessary on its blasting process – using too much energy blasting the material to create a ‘uniform’ blend. But, in upping the amount of explosive used, it created sub-optimal crusher feed.

This saw the primary crusher assigned to treat material around 5 in (127 mm) in size attempting to ‘crush’ material that was averaging around 1 in in size, according to Putt.

The primary crushing process was ineffective to say the least.

By adapting the blasting process to target the designed-for primary crush size, reorienting the mine plan so not as much skarn material was being fed into the coarse ore stockpile at once, and adding steel ball media to the mill to deal with skarn that was fed into it, the headline savings were made, according to Putt.

Such savings come with quite a bit of due diligence work, he explains.

“It is not just about connecting disparate datasets; a tremendous amount of work goes into cleaning and contextualising the data – knowing which information is right for the project at hand and which data is not applicable,” Putt said of the MinePortal data gathering and analysis procedure.

Where other data-focused companies can clean datasets and put them into algorithms to form various predictions, DataCloud’s mining knowledge and deep collaboration with customers enables the company to create fit-for-purpose solutions that work in a practical sense on the mine site.

This process sees at least six months of relevant data required up front. Then, a four-week deep dive of this data is needed to find out if the existing dataset can solve production bottleneck issues. The US-based company normally then allocates another three months to kick off the solution, on-board all teams and see improvements come through, according to Putt.

“I wouldn’t say it is a complete customisation, but there does tend to be differences in place at every mine site we visit that means the MinePortal solutions are somewhat unique,” Putt said.

Coming back to the coarse ore stockpile example, Putt recommends hard-rock miners add another filter to their existing blending process to help improve results.

“It is about adding a mill risk factor to an existing grade control program; getting the engineers to plan the mining regime in a certain way to effectively prepare the mill for the material being fed into the coarse ore stockpile,” he said.

Miners can do this by obtaining a good idea of the time window in which the material delivered to the stockpile is entering the mill, enabling engineers to trace it back into the pit and analyse the properties that were observed – and captured – during the drill and blast process.

“This can be a tricky thing to do as the size of the stockpile is changing so often,” Putt says.

Some miners use RFID tags embedded in truck loads to get a rough idea on a weekly or monthly basis when the delivered material is finding its way into the mill, but few do this on a consistent basis.

MinePortal uses machine-learning algorithms the company has augmented for geology and mining needs to automate the process.

Using features such as dynamic time warping – which measures the similarity between two temporal sequences that may vary in speed timing differences – the platform is able to reconcile timing differences from dumping ore into a primary crusher, to sitting in a stockpile, and to when the ore goes through the rest of the mill.

Putt expands on this: “There is enough robust data within a mill’s database to run dynamic time warping, a machine-learning method, to compute the delays (of the material coming into the mill) as they change.

“We don’t need the timing of the delay to be consistent; we need the data to be recorded consistently so we can find the patterns of the delays from stage to stage. Running the data through machine learning will learn the rhythms of the stockpile and filter out inconsistencies.”

At the reconciliation stage, mining companies can pair the material signatures (rock hardness, for instance) with the results from the mill (energy draw, grind size, etc).

“Typically, we find there might be one or two specific blend types that are causing the issues,” Putt said. “From there, we can carry out real-time planning to improve the operation. We then have a feedback loop where you identify the problem feeds, change the blending over the next three months and then keep running through the process for continued improvements.”

But it all comes back to ore blending.

“The best way to handle the problem is from the ore blending point of view,” Putt said. “If you can get your ore blending to be spot on where it comes with the lowest risk of impacting the mill’s performance or availability, then the mill won’t have to do anything different (change speeds, adopt new grinding media, etc).

“You still have to dig, haul and send the material to the mill, but you are sending this material to the mill in different proportions.

“It comes with the same input costs; it just requires a bit of extra planning ahead of time to save a tonne of money in the mill.”

Nokia’s Jadoul on keeping miners safe amid COVID-19

Workplace safety is a major objective of every mining company on the planet, but with the COVID-19 pandemic, for the first time perhaps, the primary danger may simply be getting too close while talking to our fellow team members, Marc Jadoul*, Strategic Marketing Director at Nokia, says.

In the mining industry, we are going to have to adapt our business practices to accommodate the current pandemic, and we have to be better prepared for similar events in the future. The pandemic has led to a re-thinking of certain safety protocols, procedures and personal protection, and it is accelerating the adoption of recent innovations that will improve workplace safety in other ways as well.

As the world has re-opened the economy, organisations such as the Center for Disease Control (CDC) in the US and the World Health Organization (WHO) have published recommendations for how to operate manufacturing and other business operations while still practicing social distancing and other aspects of workplace safety. These include having office employees telework where possible, staggering shifts to reduce the number of workers using lunch, break and washrooms at the same time, increasing physical space between employees in the workplace, wearing masks and even downsizing operations if necessary.

Given COVID-19’s ability to be spread by individuals who do not show symptoms, it is generally acknowledged that tracking contacts will be a key way to identify those who might have been exposed to a sick employee. Knowing the cost to the business of having to shut down a facility due to illness, management will need to work with public health authorities to implement practices that allow for the quick identification of suspected contacts, allow for testing and quarantine of workers in the case of an outbreak in their operation and, in some jurisdictions, be able to show compliance with these practices.

Marc Jadoul, Strategic Marketing Director at Nokia

The technologies needed to do this are not so far away. In fact, they already exist in industries where operating environments have residual risks or require robust control measures in ways that are similar to what will be needed to protect people from contracting the virus. Some of these practices have already been implemented in mines as well as nuclear facilities and high-tech chip fabricators. With some adaptation, it is not hard to see how these technologies can be adapted more broadly to make the mine workplace of the future nearly virus-free.

From a larger safety management perspective, the ultimate goal is to create a real-time, dynamic picture of what is happening with people, assets and environmental conditions at all times – what is known as ‘situational awareness’. It is crucial for conducting forensic analysis to understand the pattern of interactions and identify possible transmission paths so as to limit exposure and trigger remediation protocols, including testing and quarantining. Much of this already exists, but simply needs to be adapted to the current outbreak.

The ultimate objective of situational awareness is having 360° visibility of people, assets, infrastructure and environmental conditions. Because what you don’t see, you can’t manage. Which is important, not only for saving lives, but also for preventing productivity losses and increasing operational efficiency.

This full digital awareness of everything going on in the workplace is the main thrust of Industry 4.0, which brings together several technology streams: low-powered IoT sensors, artificial intelligence (AI) and machine learning, edge computing and next-generation wireless connectivity. These technologies combine to allow for the automation of repetitive processes, improved efficiency of operations, preventative maintenance of assets, quality control and enhanced situational awareness.

Applying these technologies to deal with COVID-19 will help to solve many of the new workplace constraints identified above. For instance, there are types of digital smart personal protective equipment (PPE) that incorporate wearable sensors and communications devices. They communicate with the operations control centre and could be used to trace employee movements, enforce geo-fenced areas deemed too dangerous for entry, or sense environmental contaminants and warn employees who have had excessive exposure to leave the area and follow decontamination protocols.

With some small adjustments, smart PPE and wearables could be deployed in many operations to enforce safe distancing between employees, using software to digitally map out work zones. They could warn employees when they are entering crowded areas or no-go zones. They could improve safety and efficiency during mustering and evacuation. And they could also enable management to forensically track past exposure of employees to those who have tested positive for the virus.

With the ongoing spread of COVID-19, mining companies need to find ways to enforce physical distancing among miners in order to keep operations open and miners safe

If sifting through location data for all the employees in a large mine sounds like a nightmare, this is where AI comes to the rescue. Sophisticated analytics software already exists that can analyse location data to look for correlations. It isn’t much of a stretch to adapt this software to smart PPE data that tracks worker movements in the facility – as long as unions and laws allow for it. This kind of software also exists to analyse video footage from CCTV cameras. All of this analysis can be used to trace infection vectors and to re-assure health authorities that protocols are being enforced on the job site.

One of the important enablers of Industry 4.0 use cases is the existence of highly reliable, secure wireless connectivity. The key to end-to-end awareness of operations is ubiquitous connectivity. Because of privacy concerns, that connectivity should be very secure. To support video and the large amount of data that can be generated within a fully automated facility, it also has to have bandwidth capacity as well as be able to support low latency edge computing. Geo-positioning and geo-fencing services for employees and mobile machines need more precise coordinates than can be provided by GPS – and need to work underground and in-building as well as on surface.

Delivering all these essential capabilities is fortunately available with today’s 4.9G/LTE and tomorrow’s 5G industrial wireless networks. Early generation wireless technologies, such as Wi-Fi, were designed for connectivity to best-effort networks. They are not highly reliable, secure or capable of providing mobility and geo-positioning services. Cellular-based 4G services, on the other hand, have been used in public mobile networks for a decade and have never been compromised. 5G is designed to be even more secure and has a number of features, like ultra-low latency, that are specifically intended for industrial automation use cases.

COVID-19 is likely to be a reality we have to live with for several years. If we are lucky and develop a vaccine quickly, it may be a short-term problem. But the scientists have been warning us about the possibility of pandemics of this nature for decades. This will not be the last. The good news is that the same Industry 4.0 technologies that are transforming our workplaces can be harnessed in this fight. Industrial IoT, edge computing, AI/machine learning and industrial-strength wireless networking will play a key role in ensuring the safety of our workers and our ability to come out of this crisis stronger than before.

*Marc Jadoul leads Nokia’s marketing efforts for the mining industry, working with key stakeholders across the business to evangelise digital technologies for creating safer, more efficient and productive mines

SensOre and CSIRO to clean up exploration datasets for AI algorithms

SensOre has commenced a joint project with the Commonwealth Scientific and Industrial Research Organisation (CSIRO) looking at automation and efficiency in big data cleaning and processing solutions for the mineral resource sector.

Access to high-quality datasets has, in the past, been an obstacle to applying cutting-edge predictive analytics to solve geoscience and mineral exploration problems, according to SensOre.

The joint project, which will confront this problem, will see several CSIRO data science experts embedded within SensOre over the next six months.

SensOre says it aims to become the top performing minerals targeting company in the world through the deployment of artificial intelligence and machine-learning technologies, specifically its Discriminant Predictive Targeting® workflow.

“SensOre is committed to world-leading mineral exploration research and development,” Richard Taylor, CEO and Director of SensOre, said. “This is the second time SensOre has worked with CSIRO and the engagement has led to order of magnitude improvements in our approach. Australian government support, such as the Kick Start initiative, is important for Australian technology companies looking to grow globally.”

The joint project benefits from CSIRO’s Kick Start initiative for innovative Australian start-ups, providing funding support and access to CSIRO’s research expertise. The program offers eligible businesses matched funding of up to A$50,000 ($34,681) to undertake research activities.

Eclipse Mining’s SourceOne can help miners prepare for the unexpected

While all businesses should have a risk plan, the very nature of mining presents a unique set of problems and opportunities for consideration, according to Abinash Moharana of Eclipse Mining Technologies.

Whether an operation is contained in one location or spread throughout the world, strategic plans impact physical activity, which, again, impacts the strategic plan. The optimisation of this continuing cycle is imperative to success, Moharana says.

Even when physical operations are forced to shut down, data analysis and other planning must continue to meet revenue projections, and to prepare for physical operations to resume. This is increasingly relevant during today’s COVID-19-related lockdowns.

“In the event of a physical shut down, all existing plans become irrelevant,” Moharana, Technical Product Manager at Eclipse, says. “New plans are needed quickly. And the most significant risk (or opportunity) at this point, is the quality and timeliness of your data, and access to that data for employees working in remote locations.”

Eclipse’s SourceOne solution, which features a collaborative platform to connect data from different sources, and a datahub to store historical and contextual data, can help here. SourceOne renders this data serviceable for analytics and for adoption of tools, such as artificial intelligence and machine learning.

The ability to create new reactive plans and multiple scenarios is imperative to surviving shutdown events, Moharana says. “And, doing so accurately may provide an opportunity to emerge on the other side even stronger than before so that when physical operations resume, you can jump right into it without wasting time or resources.”

Moharana added: “To achieve this, you must turn your biggest risk into opportunity using high-quality data that is easily accessible with real-time updates. And SourceOne can make this happen for you.”

As a mine platform designed with multi-users in mind, SourceOne is made to host multiple remote users.

It accommodates concurrent on and off-site users, while also handling off-line users with automated merges to a clean state, according to Moharana. “A geographically disparate team can work seamlessly together, transfer data and messages as part of a workflow, always be able to work with the latest data, and know the genesis of each data.”

Moharana explained: “With all the existing mine plans rendered redundant, the mine planning engineers can start working on new mine plans with updated assumptions and requirements. The requirements may have changed, such as volume becoming an immediate priority, rather than profitability, to be able to supply the end user. While doing this, the goal is also to not deviate too far from the strategic plan.

“SourceOne maintains the complete history of each project and data. Historical plans can be used to generate the differentials between the existing plans and the strategic plans. These can be weighed against the new goals, and a sub-optimal plan with smaller recovered metal may be considered as it does not stray too far from the strategic plan.”

With so many plans being made, the chances of errors magnify, as does the need to be able to audit the results with an internal or external auditor.

SourceOne maintains a complete audit trail of the entire project, so every plan can be traced back to the assumptions that were used at each step, according to Moharana. This speeds up the process to validate the suitability of the plan to be implemented in the field.

Mine planners must realise that for their corporate management, there is a big difference between risk and uncertainty, Moharana says. Risk is something that can be measured, while uncertainty cannot. The difference lies in measurable information.

“A good Business Continuity Plan (BCP) for mine plans allows the enterprise to convert some of these inherent uncertainties into calculated risks, which then can be properly weighted by the management,” Moharana said.

Mining personnel may think that the COVID-19 pandemic is an unprecedented event, but they need to be prepared for such a unique occurrence, according to Moharana.

“This preparedness allows mines to better manage the risk by being able to make mine planning a part of your BCP and ensuring that the mine is well prepared for any disruption, rare or not.

“De-risking the mine planning process is one of the many ways SourceOne can help your organisation become more resilient and be prepared for ordinary (and extraordinary) events.”

Datarock machine learning drill core analysis tool hits major milestone

DiUS and Solve Geosolutions have leveraged PyTorch-based image analysis techniques to help automate the analysis of drill core imagery and provide greater insight that can be used to influence decision making on mine sites.

Mine sites often produce between 100 and 1,000 m of drill core per day, generating hundreds of images a week at a single drill site, according to PyTorch, an open source machine learning platform.

Historically, these images have been kept as a record of the job and a resource for geologists to refer to, rather than being used as a quantitative dataset that adds value to a mining operation.

Tapping into this rarely used data source of drill core imagery, DiUS – an Australia-based technology services organisation with a strong focus on machine learning and deep learning image segmentation analysis – joined forces with Solve Geosolutions – a mining-focused data science and machine learning consultancy – to build a machine learning-powered, cloud-based platform to automate the analysis of this drill core imagery using image segmentation technology.

Together, DiUS and Solve Geosolutions worked on applying a range of PyTorch-based image analysis techniques, including image classification, object detection and both semantic and instance segmentation, to a range of geological problems.

In particular, the team wanted to understand how different models performed in terms of training and inference speed, training requirements and overall accuracy of prediction to inform how they could be deployed in a production environment.

One model they used extensively is Mask R-CNN.

“This model can be applied to a range of segmentation tasks, however it can also demand large training datasets that are sometimes not available,” PyTorch said. “To support this, the team developed novel ways to increase the initial, often sparse training dataset through data augmentation techniques such as rotation, flipping, contrast, saturation, lighting and cropping.”

Following the initial discovery period, the team set about combining techniques to create an image processing workflow for drill core imagery. This involved developing a series of deep learning models that could process raw images into a structured format and segment the important geological information, according to PyTorch.

Their first productionised process was a metric referred to as RQD, otherwise known as rock quality designation. “RQD is a difficult and monotonous dataset to collect manually,” PyTorch said. “It’s also well suited to automation, and of high value to a mining operation. RQD is used by engineers to understand the strength of a rock and is used in the design and engineering of a mine.”

With the release of Detectron2 – a PyTorch-based computer vision library released by Facebook in October 2019 – the team made the decision to switch from the previous model implementation on TensorFlow to the next-generation platform to help improve instance segmentation tasks, PyTorch said.

The team found Detectron2 to be four times faster in training the models (using GPUs) and three times faster in inference (using CPUs) than the previous model implementation, PyTorch said.

Building the models on PyTorch-based frameworks meant the team was able to reduce valuable training time across the board. This increased the number of experiments and, as a consequence, improved model accuracy on an identical dataset. The PyTorch Dynamic Graph also made it much easier for the team to debug and investigate any issues that arose, PyTorch said.

The resultant Datarock platform is a software as a service offering that applies machine learning – image segmentation technologies – to drill core imagery and delivers information about a mineral deposit’s geology at scale, and at a resolution that’s not been previously economically viable, PyTorch says.

Since launching Datarock in 2019, the team has extended the platform to turn drill core imagery into high-quality datasets to support decision making throughout the entire mining cycle.

“The models perform optical character recognition, instance and semantic segmentation, as well as geological statistical analysis on a dataset,” PyTorch said. “This allows a geologist to inspect the model prediction and check for quantity and quality in unmatched datasets.”

Mining and exploration companies can now get consistent geological information from their rock core imagery in a matter of minutes, according to PyTorch.

“This near real-time power is enabling more intelligent decisions to be made further down the mining chain – saving time and money that can be put towards other business-critical projects – and freeing up geologists to do higher value tasks,” PyTorch said.

To date, the Datarock platform has processed more than 1 million metres of drill core images – that’s enough core to cover the distance between Sydney and Melbourne – over 800 km.