
Researchers in robotics and AI at Luleå University of Technology are conducting successful field tests in LKAB's mining environment. The drone autonomously plans a safe route for its mission to collect gas data at an inspection point.
15 May 2024
Successful field tests in LKAB mine environment – autonomous robots collaborated and performed tasks independently
Researchers in the Robotics and AI Group at Luleå University of Technology have been working closely with LKAB for three years on the project entitled Autonomous drones for underground mining operations. The project recently conducted large-scale tests where the research team’s three autonomous drone robots demonstrated their ability to collaborate to inspect a GPS-denied underground environment. This is the first time anything like this has been done in an underground mine anywhere in the world, and the tests were a huge success.
“Our close collaboration with LKAB has given us the opportunity to conduct large-scale evaluations of our research, and we are now able to show that autonomous robots can be used to increase safety in deep mines. Our large-scale field test at LKAB’s Konsuln mine means we can show for the first time how robots can autonomously collaborate on performing complex tasks in a dark, inaccessible, harsh environment where GPS (Global Positioning System) systems do not work,” says Björn Lindqvist, researcher in robotics and artificial intelligence at Luleå University of Technology and project manager of Autonomous drones for underground mining operations, which is part of the LKAB-led initiative SUM, Sustainable Underground Mining.
The research group’s collaboration with LKAB has involved close and frequent discussions where the robotics and AI group, together with experts from LKAB, have established what use case types would benefit from being tested and developed in a mine environment. This has involved working in collaboration with the mine rescue service to develop drones for the future that are capable of operating autonomously in smoke-filled areas; and with the production side to measure gases after blasting, for instance.
The recent live test taking place LKAB’s Konsuln test mine, 600 metres deep, with several levels of drifts and where LKAB mines 2 million tonnes of iron ore each year, was a full system test. This means that the whole link was tested, from the operator’s input data to several autonomous robots performing a task in collaboration with one another. Nothing like this has ever been tested in a live mine environment before.
“What sets our AI and robotics group apart from others around the world is that we go all the way from lab tests and out into the real world. That is our great expertise. Moving from method development to lab test is as big a step as moving from a lab environment to real-world testing, and we would never have been able to do tests like this without our industry collaborators”, says George Nikolakopoulos, Chair and Professor of Robotics and AI.
The large-scale evaluations of the autonomous drone robots at the LKAB mine starts with the operator using a user-friendly interface to notify the three autonomous robots that inspections of gas levels at various locations in the mine are required. The autonomous central system then calculates the most efficient way of going about the work. It considers which robot should go to which point and measure the gas level, how it should navigate its way there, and then report completion of its task. All this is done in real time and by fully autonomous robots in a previously unknown, complex 3D environment with no GPS structure.
“Each robot identifies its own location and transmits its navigation, inspection data and all the rest of it in a map of the mine to the human operator to overview. When the robots have finished their current mission, they report this and indicate that they are ready for a new task. At that point, our mission control system issues new instructions.
All navigation and localization autonomy takes place inside the robot, and it acts as an individual in collaboration with the other robots,” says Björn Lindqvist.
The successful field trials at LKAB’s test mine are being conducted by a team of researchers made up of ten or so people, each of whom is responsible for a single area of research that focuses on enabling the robot team to collaborate and get the job done. The robot task allocation is to be optimised, the robots must self-localize, avoid obstacles, plan their routes and much more.
“I am really, really proud of the entire team, the robots themselves, and every single PhD student who helped to make this field test such as success. If the entire system is to be functional, the work of several researchers all has to be successful simultaneously.”
LKAB and the Robotics and AI team at Luleå University of Technology are now discussing further collaboration.
“We researchers are keen to take a closer look at how we can use autonomous systems in rescue missions. Going forward, I would like our autonomous robots to be able to put out fires or deliver gas masks to people trapped in the mine. Ensuring that a scout robot is first on scene to let rescue leaders know what has happened before humans turn up. The aim of all this is to make the working environment in deep mines even safer,” says Björn Lindqvist.
Contact
George Nikolakopoulos
- Professor and Head of Subject
- 0920-491298
- george.nikolakopoulos@ltu.se
- George Nikolakopoulos
Björn Lindqvist
- Associate Senior Lecturer
- 0920-49
- bjorn.lindqvist@ltu.se
- Björn Lindqvist