Edge computing for mining and metals applications

Digitalisation through edge computing can unearth new value in mining and metals processing.

Today the mining and metals industries face several challenges in balancing profitability and production efficiency while managing sustainability and minimising environmental impact. It is also critical that production efficiency be achieved in conjunction with flexibility to ensure industry stays nimble and adaptable in a highly volatile world.

These industries are embracing the digital transformation journey, which includes incorporation of IIoT technologies and concepts so data can be leveraged to achieve new levels of production efficiency. Gathering, storing and analysis of production and machine data form the building blocks in this strategy.

Mining, like many other industrial sectors, generates large volumes of data. Much of the initial IIoT discourse has focused on cloud computing, which meant that large quantities of raw data would be dispatched to cloud-based data lakes, analysed and used in optimisation algorithms to drive real-time decision-making.

However, mining operations are often distributed throughout many remote locations with limited infrastructure, and this presents connectivity challenges. When factoring in cloud architecture, data costs and latency, a pure cloud computing solution may not be the best answer for a mining and metals operation.

Traditional PLCs are commonly used for real-time control, but more recently have been tapped for implementing higher level IIoT communications and analysis. This change is enabled by improved IT computing methods merging with OT platforms, resulting in highly capable edge-located IT/OT controllers. Data available at the operational edge is now easily accessible and able to be processed and transmitted, paving the way for analytics and optimisation deployments.

Process overview

Mining and metals processing operations consist of many areas with opportunities for improvements through automation and data gathering (Figure 1). Sometimes the automation and communication platforms are delivered as part of original equipment manufacturer (OEM) machinery or equipment skids, while at other times they are constructed in the field.

PLCs play an important role in almost every area and may include local HMIs. Remote terminal units (RTUs) are also widely used, providing some functionality similar to PLCs and adding remote connectivity features. Increasingly, systems incorporate intelligent field devices like variable frequency drives (VFDs) that can supply extensive operational and diagnostic data. Larger processing areas may rely on a distributed control system (DCS), and plant-wide operations may be monitored by a SCADA system.

The following is a summary of typical operations:

• Extraction: Pit and underground tunnel equipment and conveyors, characterised by deployment across large areas.
• Material handling: Crushers, stackers, screeners and autonomous vehicles.
• Comminution: First stage semi-autogenous grinding mills and secondary stage ball mills.
• Separation: Includes flotation cells, leaching, thickeners, solvent extraction, and filtration.
• Refining: Systems such as electrowinning, electrolysis and smelting for extracting and purifying metals.
• Logistics: Rail and port operations for handling mining production.
• Utilities: Includes electrical power distribution and monitoring to support other systems.
• Water management: Systems to supply, treat, store, recover and reuse water to provide the right quantity and quality needed for processing.

Figure 1: Many aspects of mining operations require some level of automation, and nearly every area creates data useful for analysis and operation. For a larger image click here.

Because automation platforms and mining equipment have generally become more intelligent over the years, there are more opportunities than ever for obtaining the right field data and acting upon it to overcome operational challenges.

Edge control platforms surmount challenges

It has been possible for many years to stitch together various traditional automation technologies with satellite and radio communications, achieving some level of automated control and remote visibility. However, these solutions have usually been costly and difficult to create, operate and support.

This has changed with the introduction of a new class of automation platform called an edge controller. An edge controller combines a real-time operating system (RTOS) with a general-purpose operating system (OS) like Linux (Figure 2).

The RTOS provides direct deterministic control and monitoring of field equipment, much like a PLC or RTU. In fact, edge controllers can be used just like PLCs, even if users do not immediately take advantage of additional features, providing a future-proof design.

The general-purpose guest OS enables capabilities such as advanced computing, analytics and data storage. In addition, the general-purpose OS offers much more capable communication options, even over the low-bandwidth connections commonly encountered with mining operations.

Because the RTOS and general-purpose OS are virtualised at the hardware level on board the edge controller, they are completely independent from both a hardware and software standpoint. In fact, each OS can be independently acted upon and rebooted. However, the two OSs can communicate with each other securely using industry-standard OPC UA connectivity. This configuration preserves the robust always-on RTOS operation while enabling modern computing capabilities.

Having an onboard general-purpose OS offers the following computing advantages:

• Security: Includes defences suitable for prevalent IT-like issues like network storms and denial-of-service attacks.
• OT connectivity: Natively supports OT-oriented communication protocols, such as Profinet and Modbus/TCP, as well as newer protocols with built-in security like OPC UA.
• IT connectivity: Natively supports IT-oriented communication protocols with built-in security such as MQTT and secure sockets (HTTPS, SSL, FTPS), providing appropriately secure communications performance.
• Flexibility: Users can develop applications in IT-oriented languages like C, C++, Python, Java, and many more.

Figure 2: Edge controllers seamlessly and securely coordinate deterministic control with general-purpose computing on one platform.

OT personnel can maintain a focus on the deterministic portion of the system they are most familiar with, and IT personnel can work with the general-purpose system. The two groups can work completely in parallel, or they can coordinate and crossover as needed or desired in a clearly defined manner. Remote connectivity makes edge controllers a natural fit for mining operations, where limited staff may need to support assets distributed over large work sites or anywhere in the world.

Edge controllers can be integrated into existing systems to provide new capabilities without disrupting proven operation, or they can provide a complete control, monitoring and analytical solution for new installations.

Mining applications with edge control solutions

There are a few specific mining applications that can benefit from edge control: ore tracking, leveraging vibration data, maintenance and route planning for autonomous haulage vehicles, and power optimisation.

Ore tracking

Managing ore characteristics and variability is a key challenge after it is extracted from the ground because optimal downstream processing is very much dependent on ore size and quality. The ability to track ore size as it progresses through processing helps mine operators identify conditions that could potentially cause downstream blockages or mechanical breakdown of equipment, resulting in hours of unplanned downtime.

By leveraging a combination of sensors, instrumentation and edge controllers, the ore loading on conveyors can be analysed. Live parameters like conveyor belt tension can be monitored and analytically compared with historic data. Potential deviations can be quickly identified, and the control system can intervene to prevent a downtime incident.

Ore quality also impacts downstream processing. Tracking ore characteristics enables the mine operator to better understand energy consumption, plant feed rates and ore losses, helping them to fine-tune process parameters and optimise the overall profitability per ton of ore processed.

Tracking ore quality is perhaps the most crucial aspect of managing mine processing plants. The mineral concentration varies in the ore deposit itself. After extraction, ore is sampled and results regarding ore size, fines content, moisture level and more can be transmitted to the edge controller.

While this sampled load of ore is being transported to the processing plant by haulage vehicles, the edge controller can use the raw data regarding ore characteristics to perform analysis through embedded algorithms, and then use the results to determine the most beneficial settings for improved control of the processing plant in areas such as crushing and mill performance, and reagent dosing.

Because of the unique location and performance of an edge controller, it can take multiple inputs, perform analyses, and immediately respond to and act on changes in ore characteristics, without the latency challenges of running analytics in a cloud-based solution and then trying to implement the associated operational changes in a timely manner.

Leveraging vibration data

Mining is a mechanically intensive process with many pieces of rotating equipment. Vibration is inevitably a problem, but if it is detected and analysed it can be used to predict impending operational issues.

Equipment such as conveyors, crushers, hydrocyclones, stacker reclaimers, ship loaders and mills are prone to failure due to vibration. At the machine automation level, it is possible to collect a huge amount of data to measure vibration, temperature and noise via sensors and instrumentation.

The raw data itself is too excessive for transmission to the cloud due to costs and bandwidth issues. A better option is to relay this machine data to an edge controller, which can then use the raw data inputs to complete some preconfigured computations and send the essential time-series data to a supervisory system. This concentrated information is much more suitable for transmitting via low-bandwidth networks.

The operations team, wherever they are located, can then apply analytics to this time-series data to establish baseline equipment profiles and assess if any part of the system is becoming mechanically compromised. With this information, the team can safely bring systems offline before an adverse breakdown occurs.

Autonomous haulage vehicles

Autonomous haulage vehicles are becoming widely adopted in the mining industry, and it is critical to manage diesel consumption and optimise routes to manage their overall operating costs. Onboard controllers allow for sophisticated vehicle control with features like anti-collision and position monitoring; however, there is no embedded PLC and the vehicles do not have permanent connectivity to a plant network.

Mine operators can get valuable insight from monitoring fuel pump pulses and various other raw data which can be obtained via a controller area network (CAN) bus. CANbus is an industry standard protocol used with many types of vehicles.

In terms of route planning, the haulage truck routes can be organised in conjunction with other mechanical equipment like excavators and the crusher. In situations where a mine excavator goes offline, an associated edge controller can then communicate with haulage trucks via wireless to reroute these vehicles to operational excavators and avoid situations where the crusher remains idle. On the other hand, routes for multiple trucks require careful scheduling to avoid situations where too many haulage vehicles are queueing to offload ore into the crusher plant. This kind of live route optimisation can deliver significant savings for mine operators.

Additionally, raw data paired with analytics delivered by an edge controller can provide rich insights into the integrity of the vehicle, supporting preventative maintenance. The edge controller can be used to analyse hydraulic temperatures and other vehicle operating conditions during off-loads into the crusher plant, and then relay this time-series data back to the centralised control centre for detailed analysis.

Optimisation of power consumption

In many instances, due to the remoteness of operations, mine operators must build and manage their own power infrastructure. Much of the existing power infrastructure is based on diesel generators. With the ongoing pressure for environmental sustainability, mine operators are being challenged to reduce emissions associated with diesel generators.

Mining power plants typically run with a large spinning reserve, but there is often little visibility into active plant operations, specifically when energy-intensive equipment is coming online. When there is a lack of coordination for equipment going from standby to duty mode, sites are likely to experience brownouts due to power surges.

In the case of a standalone power plant, diesel generators are well-equipped with instrumentation and sensors, and this data can be sent to an edge controller via wireless connection. The edge controller can also analyse other raw data from the processing plant to identify when more power will be needed. Armed with this information and connectivity, the edge controller can make step changes to the power plant through its real-time operating system.

In cases where the power plant is co-located with mining operations, power plant data can be sent to the edge controller through the plant network for effective optimisation. As mine operators continue to explore renewable energy–diesel hybrid solutions, edge controllers will play a critical role in optimising overall power consumption.

Conclusion

Mining and metals operations are much more than low-tech digging ventures. These industries are looking for any technological advantage to help them efficiently and cost-effectively adapt to production and market changes in a sustainable and environmental manner.

Edge controllers are a modern automation platform for enabling digital transformation and effectively applying IIoT concepts. Because mining operations are most often remotely located, any useful technologies must be suitable for these conditions and provide extensive communication options. Edge controllers are built for this OT environment, and have the latest and most secure IT computing and networking features. Edge controllers are especially compelling for this service because they can gather and store data locally, process and analyse it, directly inform operational logic of optimal settings, and relay the most essential information to higher-level systems.

Top image: iStock.com/Dazman