Drones are becoming more and more popular in mining, but using them in smart and novel ways will make this technology truly transformative
For many of us working in or with the minerals industry, the benefits of drones seem self-evident. The technology offers tremendous gains in terms of safety, cost efficiency and, above all, the ability to do far more with our limited time. Drones can also be used in almost every aspect of the mining value chain, from discovery through to production, ore processing, transport, rehabilitation, mapping cultural heritage and even community engagement. On the face of it, the technology is more than capable of joining the suite of innovations that our industry requires in its quest to optimise and remain a world leader.
For others, however, there may be suspicion and a healthy dose of cynicism. If you are familiar with the Gartner hype cycle, you will know that there are trends associated with emerging technology that include a period of inflated expectations, often followed by disillusionment.
To address this cynicism and possible disillusionment, in this article I will outline some of the ways drones are already being used, before discussing the top three developments that will make them truly transformative for our industry.
Drone technology and uses in mining
Drones come in many shapes and sizes. There are multirotor types, fixed wing types and an exciting hybrid between the two known as vertical takeoff and landing aircraft (VTOL). Each type of drone has benefits and drawbacks. For example, multirotors have tremendous control and can more easily survey vertical faces or confined spaces. They don’t need a landing strip either. Conversely, fixed wings cannot easily survey confined or other unusual spaces but have excellent duration and can cover larger areas.
Whatever type you might be using, it should be said that one of the most important aspects in the emergence of drone technology is not actually about the drone itself – it is about what they carry. You could be forgiven for thinking that drones have literally taken off in the last few years because of their affordability and ease of use. But the one change that has made drones so successful is the invention and miniaturisation of the humble digital camera – especially cameras like the GoPro.
With this development, we now use drones to collect high definition still images and video footage. When coupled with advances in photogrammetry, we can now not only easily collect aerial footage but also create 3D models of sites, pits and infrastructure. This is far cheaper and, most importantly, safer than using manned aircraft for survey work. Using drones for this work is less time-consuming and more cost-effective than using ground-based lasers to return 3D models. It also places the capture of aerial footage and generation of 3D models into the hands of everyday workers, rather than locked away with specialists. One example familiar to most of us is the measurement of stockpile volumes at an accuracy that was never achievable using previous methods. Many survey departments around Australia have bought a DJI Phantom multirotor (or an equivalent) for this purpose.
We are of course able to do far more. A few companies, particularly the larger ones, have realised that the value offered by drones goes beyond the remit of the survey team. As an example, for the exploration or environmental geologist, there are now drones available that are so small they fit into a backpack alongside your lunch. This saves enormous time even if simply being used to help identify access in remote areas or to hunt down hidden outcrop.
As another example, visual inspections of processing facilities often only happen during a plant shut down. Drones allow inspections to be carried out while the plants remain online. This vastly reduces the need to work at height. Environmental baseline footage, inspection of tailing dam walls and pit wall imaging can all easily be captured. Drones are also being used to improve onsite road safety by monitoring traffic, berms, road degradation and hazards. In a recent article, BHP estimated that savings at their sites in Queensland were in the order of A$5 million a year (Knox, 2017).
Getting the most from your drone
To get the most from your drone, there are a few ‘tricks of the trade’ and this is where short courses or professional training from an institute like Monash University can come in handy. Highly accurate 3D models generated from photographs benefit from integrating ground control points, measured with a high precision GPS. Turning this data into information requires a good working knowledge of appropriate software. Monash has spent some years developing new tools that quickly convert 3D models of landscape and open pits into geological maps, including measurements of the orientations of faults and fractures.
Notwithstanding this, even in those companies that are applying drones in a number of different areas, it would be a mistake to think we are getting the absolute best out of drone technology. So here are my top three developments that will see drones become an essential part of the resources industry, from exploration all the way through to rehabilitation. Although there is room for the drones themselves to improve in areas such as flight duration, swarm dynamics and self-navigation, the top developments are all about the sensors and how we handle the data.
1. Sensor miniaturisation
Not surprisingly, given that digital camera miniaturisation has had such a profound impact on drones, the miniaturisation of other sensor types will transform multiple aspects of our industry. Thermal and multispectral cameras have been with us for a few years but are yet to find regular uptake in our industry. These sensors identify hotspots in infrastructure such as solar panels, pipes and flotation tanks, and monitor the health of plants and regrowth at rehabilitation sites. At Century Mine in northern Queensland, thermal cameras were useful for identifying where pyrite oxidation was occurring in subsurface rocks. These cameras can provide vital information for managing slope stability and monitoring underground fire.
Magnetometers too have been on fixed wing drones for a few years. However, we are now integrating them into multirotor machines. This will allow the technology to be used accurately in rugged terrain. It also extends aeromagnetics outside the purview of exploration. This development means potential field geophysics could be applied cheaply and quickly to open pit environments, where aeromagnetic signals at high resolution could search for near surface ore for rapid resource updates or identify voids and buried equipment.
At Monash we have drone-based laser scanning equipment (LiDAR) and hyperspectral cameras that have just emerged onto the market in the visible near and short wave infrared ranges. LiDAR by drone delivers high resolution 3D models or topographic information, even in highly vegetated areas. This technology is useful for site inspection, geological mapping, identifying drainage or monitoring biomass health in areas of rehabilitation. Hyperspectral cameras, on the other hand, help map mineralogy, allowing us to sniff out hidden resources in regions of exploration. These are not cheap sensors, but now that miniaturisation and mounting on a drone has been solved, the technology should advance quickly and bring costs down.
And what of the near future? Our imagination is not the limiting factor. The trick is industry-focused investment. At Monash we have projects under development in gravity and potentially even seismometer deployment. Other sensors being discussed around Australia include ground penetrating radar or electromagnetics. If we want to see instruments that are robust and tailored for our industry, being the ‘best followers’ will likely only result in us waiting five years for instruments that are not quite up to scratch.
2. Real-time information
No matter what type of sensor is used, it does not take long for large volumes of data to be generated. For the resources industry, the most relevant issues are how to use this data to its full value and how to do so quickly. In the case of photogrammetry, LiDAR, hyperspectral analysis etc, the data needs to also be processed before it can be turned into useful information like geology maps or measurements of ground disturbance.
The solutions to these problems already exist. But we must adapt them to our needs. Part of the solution is machine learning, using graphics cards on board, and then processing the data.
The second part is transmitting the data to a laptop or even high-performance computer. From here we use algorithms to turn the processed data into maps, measure features, or flag areas of infrastructure degradation. This is the process of turning data into information while the drone flies and it could make our jobs much easier.
The third point is related to the fact that, according to the hype cycle, emergent technology can often become associated with disillusionment. One outcome of this can be ‘underutilised uptake’, meaning technology that is integrated into operations but is not used to its full potential. In contrast, interoperability is the capacity for technology to be used for multiple roles at once. A simple example would be attaching temperature and fume sensors to any drone being used to photograph and survey pits. Fume sensors are inexpensive and can be less than 50 grams in weight. Likewise, the exploration geologist using drones in the field to identify or map outcrop is in fact also collecting environmental baseline data at the same time. In this case it is important that the geology and environmental divisions of a company talk to each other and share data.
A few companies have recently developed drones that can fly autonomously underground while scanning with a laser. These return 3D models of drives and stopes. This is something we have also been exploring at Monash. In our case though, we are combining lasers and photogrammetry alongside real-time object identification and mapping. With such a drone it is possible to survey stopes at the same time as generating geology face maps, inspecting rockbolt integrity and measuring tunnel convergence. Such a tool accelerates the time that geologists, ground control engineers and surveyors require to do their job. Instead of spending many hours travelling underground, visits can become more focused and armed with much more accurate knowledge about the orebody. This also is a good example of using drones to achieve something we have never been able to do before – which is to return geology information from inside a stope.
The incredible promise of drone technology is its ease of use. The dominant limitation is not the technology but regulation and industry-specific compliance. However, this too is changing rapidly. We have before ourselves an opportunity to take many techniques out of the hands of specialist operators and integrate them into the day-to-day workflow. It is for these reasons we should be looking for sensor and data analytical solutions that can be mounted onto any cheap, off-the-shelf, commercial drone and flown the next day. The real question to get us all thinking is: what can this technology enable us to do that was never possible before?
Monash at IMARC 2018
Steven Micklethwaite will be presenting on ‘The new possible: Drones across the mining value chain’ at IMARC 2018. Other Monash staff including Profs Jacinta Elston, Elizabeth Croft and Edward Buckingham will also be sharing their insights. Monash University will also be in the IMARC exhibition at booth E50. Visit imarcmelbourne.com for more information.
Knox F, 2017. How drones are changing mining [online]. Available from: https://www.bhp.com/media-and-insights/prospects/2017/04/how-drones-are-changing-mining