There’s more than meets the eye
Intelligent machine vision system applications in mining
Written by Gurjeet Bansal, Corporate Communications Manager, IntelliView Technologies Inc.
Mining operations normally run for 24 hours a day, year after year. An unplanned downtime – whether due to an environmental concern, a safety issue or equipment failure – could result in significant output deficits, financial losses, and prolonged business disruptions.
Given the unfavorable impacts of operation derailments, it is no surprise that a visit to a mine site will reveal the complex approaches employed for process condition monitoring. There would be workers on site undertaking quality control protocols, security personnel and cameras on guard securing the premises, and other process specific technologies.
In this safety-conscious and eco-driven climate, companies are propelled to regularly evaluate their systems to ensure compliance with industry and their own standards. New technologies are continually being developed to meet the momentum of evolving business challenges amidst goals of raising operation productivity while protecting workers and the environment.
Evolution of Visual Technologies
Along with personnel, Analog/IP cameras and thermal imagers account for the most common forms of industrial visual surveillance methods. Historically, these have been employed independently and served different purposes albeit with some degree of overlap.
CCTV cameras have been a staple for many decades now for security surveillance, which in recent years received an update in the form of IP or network-based systems with high definition resolutions and motion detection capabilities.
Meanwhile, though adoption has been slow, thermography is increasingly gaining validation as an inspection and predictive maintenance tool. Impervious to ambient lighting and most weather conditions, the technology reveals heat signatures of objects, including temperature abnormalities that signal the likelihood or the presence of a problem.
Handheld and drone-mounted thermal cameras are the most common formats of thermal imaging used today, owing to the portability offered by their compact and lightweight form. These can be hand carried or flown by an operator to the different areas of a mine to inspect for potential hazards and anomalies such as machinery overheat, fluid or gas leaks, equipment wears and other inefficiencies. Its small footprint also allows the camera to be easily maneuvered and pointed at hard-to-reach spots while maintaining a safe distance.
However, like other manually controlled devices, operation is limited by personnel availability and associated costs. This introduces gaps in monitoring that can lead to issues being missed, particularly at unmanned and sensitive sites that require continuous monitoring. Training and data interpretation error are additional concerns with such technologies.
Technology Convergence – Getting Smart and Getting It Right
An evolutionary advancement that emerged from existing visual technologies is the convergence of optical sensing, thermal sensing and video analytics packaged into a single product, such as the Dual Camera Analytic Module (DCAM™) developed by leading Canadian software developer IntelliView Technologies Inc.
Thermal sensors form heat zone or infrared radiation-based images, as such they are unaffected by lighting (or the lack of), rain, fog, snow and environmental conditions. High definition video cameras produce vivid images that can even reveal details of small objects and faces – the perfect verification tool for making sense of a thermal image.
Artificial intelligence allows various degrees of analysis and decision making to take place independent of human intervention. In visual surveillance applications, it automates detection and notification of events based on set parameters – saving hours of video footage viewing time and reducing periodic and event-triggered site visits. Control center personnel can simply be alerted to pre-qualified events within seconds, visually validate events from their current location and make immediate decisions.
Implemented in a fixed platform, this system configuration delivers uninterrupted monitoring. The likelihood of missing an event is much lower than with handheld solutions. Given its industrial internet of things (IIOT) architecture, the analytics software can be applied at the edge, in the cloud and/or offsite. Processing video data in real time at the site minimizes bandwidth consumption and any potential delay in reporting.
Systems such as IntelliView’s intelligent vision systems, which are driven by built-in patented and proprietary image processing software, include video clips in their notifications and provide web access to onsite system settings, live video and event data.
Reduced operator dependence, increased monitoring efficiency, offsite detection and validation and informed decision-making that lead to effective risk mitigation are some of the other advantages of this technology progression. Industries are gearing towards full process automation to increase operation efficiencies, safety, and productivity. The broadened capabilities and applicability of thermographic imaging in mining environments brought to the fore by artificial intelligence is certainly welcome news.
With perimeter security being paramount to mine operators, the employment of a combination of observation techniques – from CCTV cameras to guards and patrolling personnel – is typical. In nonstop operations, relying on these methods alone can be rather human intensive and not cost effective.
Enhancing site surveillance with intelligent machine vision systems provides not only constant monitoring but also significant reduction in man hour fees. These systems are capable of detecting motion, intrusion, loitering, theft, objects left behind, and other security breaches.
AI today has surpassed the motion detection era to provide face detection, object identification and other intelligent detecting capabilities. Automated machine vision systems can also be linked to access control systems for guarding unmanned gates and restricted areas as well as regulating people and vehicles entering and leaving.
Detecting Leaks Early
For metals extraction companies that utilize and transport hazardous gases and liquids, mitigating leak-related environmental, safety and financial risks is top priority. In this endeavor, early leak detection plays a crucial role.
Traditional pipeline leak detection methods include flow/volume measurement, pressure monitors, acoustic sensors, and manned patrols. Thermal sensors, including optical gas imagers, are a newer technology commonly deployed as drone-mounted, handheld or fixed systems. Despite the availability of these options, some coverage inadequacies still remain due to the technology’s cost-prohibitiveness, unsuitability, and/or non-continuous running time.
A fixed thermal/optical system with leak analyzing software overcomes most of these constraints through continuous monitoring and automatic detection and alerting, and can replace or enhance existing systems. The role of leak analytics is to check for parameters, such as temperature, attributes and behavior, to determine if a leak has occurred. Gold, copper and related mine facilities operating leach pads, for example, have started adopting this new methodology to detect cyanide, sulfuric acid, and slurry leaks.
Equipment Overheat Detection and Fire Prevention
Frictional heat can develop from operation of certain machineries and handling of hot materials through mining process cycles. Conveyor belts, for instance, are known to overheat and trigger fires, and may have to meet fire resistance standards. Moreover, downtimes to repair are long.
Round-the-clock surveillance is recommended for such combustible environments. Ruggedized or explosion-proof analytic thermal/optical imagers can help avoid fires and equipment damage by detecting critical temperature thresholds in real time and generating alerts. An early warning allows workers to intervene and implement remediations, consequently preventing a full shutdown. The technique can also be deployed for monitoring other heat generating elements, such as crushers, motors, and bearings.
Mechanical/Machinery Early Failure Detection
Temperature can indicate the health condition of equipment. Routine physical inspections using mobile thermal cameras are often used for predictive maintenance and to periodically check for existing electrical or mechanical issues. Machine vision systems with temperature detection software built in are a good alternative for un-manned or remote environments, where ongoing monitoring is needed.
In the case of a mining substation, this solution can alert on a failing voltage regulator. An advance notice gives the operator sufficient time to order a replacement and book an electrician before the component actually fails and shuts down power.
Addressing Mining Challenges and Digitalization Needs
Various challenges of process condition monitoring and compliance (safety and environmental) in the mining industry can be effectively addressed with dual (thermal/optical) imaging systems powered by artificial intelligence. Although still new to mining, the technology combination has been rigorously tested and accepted in the oil and gas industry. It has cut the process of detection to alerting to mere seconds, allowing decisions to be made quickly. For unsafe situations that could have disastrous and costly repercussions, such as spill-related environmental damage or overheat-triggered explosions, prevention and early detection are key.
A proactive industry like mining is aware of the value of the right technology. However, adopting a new technology also puts discussions about balancing safety and environmental goals with productivity and bottom line projections on the management’s agenda.
In this age of digitalization and automation, technologies are expected to do more than just deliver on their technical capabilities, business advantages and extended benefits. It is now also important for technologies to be able to integrate and communicate with existing systems across the different departments of a mining business and be able to accommodate the changing needs of a growing company.
IIOT compliant AI solutions address these accessibility, scalability and interoperability requirements. However, when choosing a provider, their software’s capabilities, detection accuracy, false alerts performance, reliability and optimizability are key things to consider. Other important considerations are a company that develops its own software, has technology patents, and has a reputation backed by major industry players.
About the Author
Gurjeet Bansal is the Corporate Communications Manager for Canadian company IntelliView Technologies Inc., a leader in video analytics and AI-driven machine vision systems designed to cost-effectively improve operation efficiency and mitigate financial, safety, and environmental risks through the automation of real-time leak detection and industrial surveillance operations. Gurjeet has more than 15 years of combined experience in the field of marketing communications within media, agency, and corporate environments. Her educational background consists of a master’s degree in integrated marketing communications from the University of Westminster (England), a master’s degree in event management from the University of Technology Sydney (Australia), and a bachelor’s degree in journalism from the University of Santo Tomas (Philippines).