Applications of cloud computing in the mining sector

Concept:

Colorado's Newmont Corporation (Newmont) has collaborated with a US-based GIS mapping software provider Esri to use the latter’s cloud-based Site Scan for ArcGIS platform to supply drone workflow tools under its Global Drone Operating Model. Newmont Corporations intends to standardise and scale its drone operations. Newmont has implemented the Site Scan for ArcGIS at Boddington, Australia, and now aims to deploy it at the Peñasquito Mine in Mexico.

Nature of disruption:

Newmont utilises Site Scan for ArcGIS, along with ArcGIS Pro for reclamation reporting, improvement of health and safety initiatives, and monitoring and analysis. Based on the use of end-to-end cloud-based drone mapping software under the Global Drone Operating Model, Newmont has developed standardised procedures, wherein all drones create high-quality 2D and 3D imagery to visualise and analyse the output on the web. It enables information sharing, publishing the data within ArcGIS Online or ArcGIS Enterprise, collecting and processing files using reality capture technology within Autodesk BIM 360 or exporting them in common file formats.  

In addition, a Site Scan for ArcGIS can be used to examine topography and slope stability for improved protection at Newmont mine sites, as well as to audit inventory of mine and area changes. The end-to-end solution will allow Newmont operators to promote decision-making over the life cycle of mining, manage their drone fleet, and drive productivity.  

Mining operations are performed in remote locations, making it time-consuming and expensive to track mines and associated activities through conventional methods. It requires a large workforce to observe the operations, which raises safety concerns. Besides, Newmont’s unorganised operations hindered the potential of drone management software. The deployment of Site Scan for ArcGIS has helped Newmont to facilitate monitoring, mapping, and surveying of remote locations, which resulted in improved decision-making.

Concept:

Personnel fatigue is detrimental to costs and safety in heavy industries. Predictive fatigue management works according to similar principles to the predictive maintenance many companies now use for asset health. Necessary interventions can be anticipated by extrapolating from collected performance data, and pre-emptive action can be taken before any negative consequences manifest.  

For example, rather than a traditional reactive approach like shaking a driver’s seat if he has fallen asleep, a predictive approach might organise task assignments so that workers that have not slept for more than a certain number of hours cannot operate heavy machinery.

Nature of disruption:

Fatigue Science’s Readi Fatigue Management Information System (FMIS) monitors personnel fatigue, analyses the data, and enables smarter task assignment and smarter dispatch for optimum productivity and safety. Other features include recommendations on improving sleep conditions and real-time fatigue alerts. Sleep data is captured from wearable devices. Readi’s machine learning engine analyses the workers' sleep and work hours over the previous ten days.  

The transition to the cloud enhanced Readi’s effectiveness significantly. Previously data would be gathered, and a report would be presented to the client up to four times a year. Now automated reports are provided every day at the start of the shift. Also, the data is available to everyone in the company. 

The two mines in question saw two types of benefits from Readi FMIS: the first is fatigue reduction, which led to workers performing tasks to a higher standard; and the second is task assignment optimisation, which meant that workers were assigned the tasks that best suited them.  

At the two mines in question, it was found that the first benefit brought $3.3m in output gains and the second $2.3m. Savings derived from the reduction of fatigue-related lost time incidents brought the total gain to $5.9m.

Nature of disruption:

Metso Outotec provides equipment and machinery to industrial companies. Its customers demand effective and reliable products. In hopper, crusher, screener, and conveyer development, Metso Outotec previously used Ansys, which simulated various structural conditions to test the machinery. In these simulations, rock load behaviour was troublesome.  

The results often seemed suspect, but there was no way to assess the process because the simulations were black boxes: input was entered, output was returned, but what calculations occurred in between was opaque. Inaccuracies meant designs had to be iterated and tested multiple times, which cost money and time.

Outlook:

Metso Outotec decided to introduce Rocky DEM software to their Ansys simulations so they could visualise and observe the simulated behaviour of the rock loads. Once calculations could be validated or corrected, the modelling improved, and far fewer prototypes were needed before a product was ready for market. Rocky DEM uses computing resources heavily.  

Generally, such resource-intensive programs would be run on high-performance computing (HPC) systems. Metso Outotec had some in-house computational capacity, but more would have been helpful. Engineers elected to outsource the computing to the cloud rather than install more capacity, since the extra power would only be necessary when Rocky DEM was in use.  

Rescale is a platform that lets customers access and use HPC in the cloud. Rescale facilitates access to Microsoft Azure, Google Cloud, and AWS. Metso Outotec enlisted Rescale for its Rocky DEM simulations. With Rescale and the cloud, it was possible to do many more simulations. Fewer hours were required, and costs went down significantly.

Concept:

Newcrest operates projects in six provinces across four countries, focusing on the Cadia East (Australia) and Lihir MOPU (Papua New Guinea) expansion projects, as well as the Wafi Golpu (Papua New Guinea) and Namosi (Fiji) greenfield projects. 

A challenge faced by Newcrest is keeping mining machines running and maintaining continuous production of gold at Cadia Valley, Australia's largest underground mine. A key element of this process is regulating crushed ore levels in the ore bin, which sits downstream from the crushers. Maintaining the correct level is vital. An under-filled ore bin would lead to ineffective filling of collection conveyors, while an overfilled bin would need to be emptied manually, causing safety issues.  

Originally, hard sensors used for monitoring were unreliable in the challenging underground environment and failed two to three times per month, forcing the company to use visual inspection. The frequency of failures led to 4,780 minutes of downtime over six months and significantly increased safety risks with manual repairs needed.

Nature of disruption:

Newcrest partnered with Insight, an IT services provider, to integrate Microsoft Azure's IoT capabilities, collect data from sources upstream, and predict the level of crushed ore at any point. The software improves the flow of the mineral in the production process by improving the regulation of ore levels in the crusher.  

The new solution has an 85% accuracy in predicting crushed ore levels, and the production of ore can continue for four hours after a hard sensor has failed. Using the cloud as a development platform, a new application was built and deployed in days rather than months.