Visual inspections of underground mines are known to be risky, time-consuming, and expensive – but drones prove they do not have to be. By Siobhan Doyle.
With the world’s population rising, coupled with increasing demand in resources, the mining industry is under pressure to supply more minerals to meet such needs and expectations. This has resulted in mining operators extracting these minerals from greater depths.
But delving deep underground to mine comes with its many challenges: the deeper the mine, the more adverse the environment and the larger risk of rock fracturing, prompting companies to provide solutions to keep miners safe.
To put an end to hazardous inspections traditionally carried out by miners, Swiss company Flyability have developed drones that can remotely inspect, survey and map inaccessible areas from a safe distance, providing underground insights that have not been possible before.
The Elios 3
With its small form factor, Flyability’s flagship drone, the Elios 3, and its accompanying surveying payload, can fit through openings as small as 50x50cm. It has a fixed cage that physically protects the hardware and a patented combination of flight controller and motor design that allows the drone to recover flight stability after a collision. This enables the drone to navigate through the most complex environments to capture high resolution scans.
“The whole structure and body of the drone is designed to be as rugged as possible,” says Michael Blake, product manager at Flyability. “We have tested the drone in different types of environments – we even built a sauna in our basement and flew the drone inside for around 150 hours until something failed to test it in really harsh environments.”
The drone also has a modular design which features two ports: one dedicated to a LiDAR scanner and one for auxiliary payloads. This unique integrated design provides aerial capabilities ideal for indoor applications such as underground mining. Blake says the company wanted to create a modular done to “provide the same flexibility to customers” and “guarantees hardware and software updates that are likely to come in the future, making it an easier investment for them”.
Surveying the inaccessible
Comprehensive data is key to supporting operational decisions in mining, yet many hazardous environments are beyond the reach of existing data capture technologies. Blake, who has previously worked in underground mining, stresses there are a lot of areas where it is unsafe to enter because of rockfall risks, weakness in the structure of the tunnels, lack of visibility and water.
“Operators used traditional surveying methods such as terrestrial laser scanners (TLS) that they would push into these areas to try and get scanned data. But these methods are not sufficient, especially in mining environments, as there are so many areas of danger or they are locked off,” Blake explains. “But mining operators really need to gather critical data from these areas to make decisions.”
To tackle this, the team at Flyability developed a specialised LiDAR payload that comes with its third-generation drone, the Elios 3. The payload turns the indoor drone into a flying mobile scanner that can fit through openings as small as 50x50cm and create high-resolution scans beyond line of sight. This allows mining operators to collect data in areas not accessible with traditional tools and other UAV technologies.
Imaging capabilities
The drone from Flyability is powered by a SLAM (Simultaneous Localisation and Mapping) engine called FlyAware which combines computer vision, LiDAR technology and a Nvidia graphic engine. Acting as a centimetre-accurate indoor GPS, it builds real-time 3D maps enabling the drone to sense its surroundings instantaneously.
Meanwhile, the drone’s embedded LiDAR constantly scans its surroundings to build a high-density point cloud map. This map is displayed instantly in Flyability’s Cockpit App, empowering pilots with greater locational awareness. A real-time 3D view of the drone’s trajectory and environment helps mining operators obtain crucial data. Pilots can swap between the visual camera feed or navigating using the 3D point cloud, which is helpful when flying beyond the line of visual sight or in dusty environments where the camera feed is unclear.
The LiDAR scanner on the surveying payload is a OS0-128 Rev 7 sensor from LiDAR technology company Ouster which has allowed Flyability to increase the range, point density and precision of the LiDAR itself. When it comes to high-grade mapping, Blake says this sensor has enabled operators to see further away from the drone and had helped make the drone software more efficient in mapping out these sites.
“It was a no-brainer when we created the Elios 3 to have a LiDAR sensor onboard. Now the surveying payload has improved what our LiDAR is capable of. We can see more in the same flight time, enabling us to get better situational awareness from flying. The pilot can better understand the environment, the distances between areas, and inspect particular environments the drone may come across.”
The front of the drone also features a 4K video camera that enables operators to take photographs of any objects seen down in the mines – for example, a cracked roof or can see water coming in. “They can take a photograph and that image becomes a point of interest, and these are then localised on the map and are made readily available in a high-density point cloud,” Blake adds. “This allows the mining manager or operator know where this photo has been taken relative to the structure of the mine through the LiDAR map – they can easily identify and locate a defect within the site from the drone’s map.”
3D surveying solution
Flyability also offer software to accompany the Elios 3 that turns the data from the drone into survey-grade 3D models, helping surveyors map unreachable places such as mines. “After doing a flight of the area of the mine they intend to map, operators can download the raw data from the LiDAR map itself and post-process that on an application called FARO Connect developed by our partner, 3D measurement company FARO Technologies,” says Blake. “The software has SLAM presets – optimised for data gathered with the Elios 3 – that are more advances and can correct any errors within the high-grade data itself.”
During processing, FARO Connect automatically interprets reflective targets as ground control points to georeference Elios 3 point clouds, as well as filter them for noise. With these targets, it can also align multiple point clouds to create a single, clean result for clear visualisation. “By using this software, they can create a high-accurate map of the mine and reduce any of the drift factor that may have been accumulated during the flight.”
The survey-grade mapping solution also features loop closure and global optimisation which looks at all the points then distributes them to make sure all the points have been covered in the area of the map itself. “FARO’s application gives out around 1.3 million points a second,” Blake highlights. “It is a much denser point cloud which is better for making decisions and can pick up a lot more smaller features, or errors, in the map.”
Future of drones
Blake is optimistic about the use of Flyability’s drone technology in the underground mining realm, but highlights the company’s future goals, automation being one. “Although the drones are easy to fly, because mining companies change personnel all the time, you may come across someone who has never used this type of equipment without much training – we’re trying to make the whole operation use of the drone easier.” For this reason, the team at Flyability have focused on making their drone easy to use including some automated features such as Return-to-Signal.
“Flyability is working in partnership with mining companies around the world to determine what future developments for the drone and its associated software will be most beneficial for optimal use in underground mine environments.”
Blake also envisions the mining world becoming more welcoming to artificial intelligence (AI), highlighting future hopes to start running an AI algorithm on data collected by their drones to look for features such a fragmentation on the grounds.
“Mining is one of the slower industries to adopt this type of technology because they initially did not want to change their methodologies too much. However, that is changing. “We hope to be at the forefront of that in terms of using new technology such as AI in the mining scene, and to be able to harness as much of the current workflow and also bring in new workflows to make everything as seamless and efficient as possible,” Blake concludes.