Extremely large sites, multiple facilities, high volumes of pedestrian and vehicular traffic and very valuable assets are common characteristics of mining operations. Securing these behemoths is often a logistical nightmare as threats from common thieves to illegal miners are added to the inherent safety and security risks synonymous with mine complexes. Hi-Tech Security Solutions discusses solutions with industry specialists.
Charles Harrison of Secu-Systems says that in essence there has been a fundamental change from traditional perimeter intruder detection systems for mines. Typically, mining operations now tend to spend less money on installing physical barriers such as fences and have moved towards investment in fibre-based technology, wireless point-to-point, camera following camera or beams, typical intrusion solutions on critical infrastructure internally, due to the complexity of operations and the sheer distances that need to be covered.
He adds that another prompting factor towards additional technology adoption is that once an intrusion has been detected, the time taken for a reaction team to find and reach that point is often excessive, with criminals having entered the facility and further detection proving nigh impossible due to the vast area to be covered.
The trend is to institute the peeling-onion strategy from the inside out, says Harrison. This allows for 360° wide area surveillance solutions. On large sites that include opencast mines or even country borders, physical barriers do still play a role, but detection options such as high-end thermal and day/night HD camera solutions are being strategically deployed in high-site locations.
This deployment is often coupled with Australian-developed Panoptes military type Moving Target Indication software. This allows long-range ground-based surveillance sensors to cover 20 to 30 times more area with the same assets. Further, it allows for multiple target identification and tracking autonomously and simultaneously. Cameras can easily detect movement in as little as 4 pixels. This software can also be coupled with ground radar in areas that allow for it, such as fairly flat areas whereby slew-to-cue options can be used. The cost benefits are that it is the cheapest surveillance system per square kilometre available on the market.
Hope for the best
Dene Alkema of Cathexis says that thwarting the ability of criminals to ‘beat’ perimeter security systems requires a thorough assessment of existing systems and threats. “Not much time has been spent on understanding the risks encountered on mines, the threats posed by outside forces or how to address them. Traditionally, a stock standard approach is adopted, with the erection of a fence and the deployment of a couple of surveillance cameras, followed by a ‘hope for the best’ attitude.”
He continues that one needs to determine exactly what it is you are trying to protect. Is it infrastructure, resources, health and safety, or production uptime? In order to outline a comprehensive risk mitigation plan it is important to take the specific environment into consideration, carefully identifying high-risk areas. Addressing these risks is normally an exercise tempered by cost constraints and cost of ownership concerns.
Success, Alkema claims, can be achieved by identifying where physical barriers will assist with the risk plan and then adopting a layered approach to holistic security. Each mining operation will have an individualised approach in order to ensure maximised goal attainment.
Operational systems need to be able to detect and possibly identify criminal risks and apart from cameras, there are various methods of detection, says Alkema. These include fibre fence, microwave and radar technology. By integrating devices and layering technology one can increase situational awareness.
Johan Crause of Arecont Vision says that there is no doubt that a physical barrier is probably the best way of prohibiting people from entering mine premises, but unfortunately the cost of installing a long fence and constantly maintaining it generally outweigh its advantages.
Megapixels on the perimeter
Arecont, he says, focuses on supplying high-megapixel surveillance technology to provide perimeter security. This technology, because it is able to clearly identify perpetrators, results in the need for fewer cameras on the perimeter, simultaneously resulting in reduced capital outlay and maintenance and the ability to achieve a better overall view of the premises.
High-megapixel camera technology is based on mathematical calculations which use pixel density and analytics, on the backend, to identify and recognise whether an object detected is a person or animal. Using the ambient scene relayed from the camera, operators can see in what direction the detected person is moving.
This then enables the control room operators to dispatch a response team should the person not be authorised to be in that area. Crause cautions that high-megapixel technology should be used in combination with a thermal at night since it will not be able to produce a clear image in low or absent light. Similarly, a thermal camera will not produce clear identification during daytime.
Callum Wilson at ISDS says that one needs to look at overall security as a layered approach, since there is simply no silver bullet. In essence, security should always be a combination of techniques and technologies, with physical security and technology playing symbiotic roles.
In terms of perimeter protection, there are three factors under consideration: the initial barrier which should deter and stop perpetrators which should include early warning alarms such as PIDS; the surveillance aspects; and the problem that arises in terms of what happens when the criminals breach the perimeter.
“If one had limitless amounts of money, then the deployment of extensive high-end technology would be the solution, but given typical security budgets this isn’t always possible. While one should never underestimate the value of having a physical barrier such as a fence in place, technology such as surveillance cameras is deemed critical. There needs to be a balance with, in my opinion, a major proportion of the budget being allocated to camera and other technology rather than the bulk allocated for a suitable barrier,” Wilson points out.
I see you
“The question of whether movable cameras or static thermal cameras are more effective in specific areas cannot be answered simplistically. The largest challenge when using surveillance cameras is to drive down false positives due to environmental influences, for example in difficult scenes such as dense bush. Operators often become weary and jaded and ignore repetitive scenes. The adoption of both technologies will assist in detecting intruders even in inclement conditions and should be the first choice of any mining facility,” Alkema points out.
With regard to tracking cameras, Crause says that issues arise when more than one perpetrator is present in a scene. When the camera starts tracking one particular person then the other people are not being tracked. Therefore, he believes, static camera technology is preferable since you will not lose sight of the primary objective, which is to detect a person or persons in a specific scene, identify the direction in which they are moving and then dispatch a response team if required.
Wilson suggests that technology should be deployed on a variety of levels, for short, medium and long-range distances. For shorter distances, one can use static cameras which create a fixed view of the environment. These would typically be installed around compounds or sensitive/key areas where they allow for a variety of analytics to be undertaken on the fixed environment both from a biometric perspective in terms of facial and licence plate recognition as well behavioural analysis to create situational awareness.
For medium to longer distances, he says PTZ cameras allow one to interrogate the wide areas far more effectively. In areas, such as on mines, where distances to be covered are often extensive, one would require thousands of static cameras, whereas fewer PTZ cameras are needed for these distances. The use of thermal or electro optic infrared (EOIR) technology is also suggested, as key areas of ingress are often targeted at night.
Harrison says that commercial cameras can attain detection radiuses of only 2.5 km, whereas military-spec cameras, such as the FLIR HRC-X, can attain detection distances of up to 17.8 km or the FLIR PT-602CZ can attain target detections up to 9 km. This makes these cameras extremely beneficial on large sites and in ambient conditions of total darkness, smoke, dust and light fog.
Worth noting is that while thermal cameras give users the ability to see what is going on, irrespective of the level of light and in adverse weather conditions (since they rely on heat to detect movement or motion), they cannot be used for personal identification purposes. Therefore, a combination of thermal and traditional cameras is suggested as best practice.
Wilson believes that drones will have a major role to play in mine security in the future. These airborne camera vehicles allow coverage of large areas and they are quick to deploy and react. A drone is therefore the first interrogation tool at an event, providing rapid relay of visuals for advanced reaction. However, they do require a certain skills set which may be an operational burden to many organisations. This could, however, be alleviated if security companies were to offer drones as a service.
On track
Is there a way to remotely identify authorised people moving around the terrain and assure operators that they have permission to be in an area? Alkema says that one needs credentials, such as fingerprints and tags, to clearly identify the authorised person. Behavioural analysis is being considered by some operations. This allows the technology to notify the operator of unusual movement or distinct variations in an area. While it is still in the early developmental stages, as the algorithms improve it will make more definite and accurate identification more achievable.
A new, but slow-uptake trend is towards using facial recognition technology. The high costs of this technology and its previous limited applicability in terms of usability in some environments and for specific culture groups, have been barriers to adoption, but this is gradually changing. “We are personally investigating the adoption of facial recognition technology for specific projects going forward, as it becomes more affordable and usable,” he says.
Thermal tags issued to enrolled users and utilised in conjunction with thermal cameras, allow mines to differentiate friend from foe. Harrison cautions that because this is a military application, end user certification is required.
Wilson says that where people are present within the working/controlled areas of the mine such as buildings, then one could consider tracking technology other than tags, such as automated facial recognition or gait analysis, but for less confined and controlled areas this would be too costly.
Artificial intelligence (AI) is being increasingly adopted by the video management market for identification purposes. “Surveillance systems are ‘learning’ certain characteristics. For instance, let’s say that a security officer regularly patrols a specific route at exactly the same times every day, then the VMS is made aware of this behaviour, notifying the system of the officer’s accepted presence and then removing the alarm,” Crause says.
In essence AI provides learned environments, whereby it identifies and recognises threats, while at the same time excluding known acceptable parameters. It is usually a plug-in on top of the VMS system as an additional feature and is, according to Crause, worth its weight in gold as the biggest bugbear for operators is false alarms during the surveillance process.
Best practice
One of the best practices for mines is to use an open platform. “Unfortunately, many mines have closed systems which this limits their system operability. We believe that from a thermal technology and early recognition technology perspective, it is always best to isolate threats before they happen, rather than being reactive. Finally, it is advisable for mine operators to consider options other than the cheapest camera. Mine facilities are generally known for their rugged and inhospitable terrain. Cameras should therefore be manufactured and tested to the highest standards and be proven to work in these environments. On that note, system integrators need to do their homework on the latest technologies available since this is an ever-evolving market,” says Crause.
In a nutshell, says Alkema, mines need clear security policies that are measurable. Operator competency needs to be constantly checked and tested to ensure that there are no loopholes in the security chain and training needs to be perceived as a crucial part of induction and ongoing maintenance of efficiencies.
Technology, says Wilson, needs to be an asset aggregator; helping mines to get the most out of their solutions or infrastructure. The primary goal of any security system should be to, deter, detect and react/intercept and this can be enhanced by early detection, establishing where the reaction points are and how quickly do they deploy.
Return on investment through surveillance camera adoption is guaranteed. Harrison cites the case of a mine that is currently installing new cameras. “The actuaries have performed calculations on the four new cameras in terms of their cost effectiveness and in only three weeks the cameras paid for themselves twice over. This was primarily due to the fact that each week their footage led to the arrest of between 80 and 100 illegal miners. Incurred losses from illegal mining have been totally obliterated. The successes we have experienced here have led to us embarking on a drive to enter the anti-poaching arena with our technology solutions.”
An holistic, multi-layered approach that focuses predominantly on a variety of surveillance technology, combined with physical barriers and a reaction component, is considered the best approach for maximising security on mines.
