Prof Michael Milford, Mining3 Technology Leader – Automation reports “2018 has been a productive year for the Automation-Enabling Positioning for Underground Mining project, a joint initiative between the Queensland University of Technology, Caterpillar, Mining3 and the Queensland Government’s Advance Queensland Innovation Partnerships scheme.
“The project is developing a suite of positioning technologies for underground mine site operations: a tracking system that uses cameras and provides the approximate position (to a few metres) of moving assets in the underground mine site, and a precise multi-sensor-based positioning system that is accurate enough to enable autonomous vehicle operation..
“In close collaboration with our industry partner, Caterpillar, developments this year include the maturation of the camera-based positioning system and the refinement of several iterations of the sensor rigs that are attached to the vehicles. We have now performed field experiments gathering datasets on both light vehicles and an actual Load Haul Dump (LHD) machine—one of the primary envisaged targets of the technology—across a number of mine sites in Queensland and New South Wales. Our research engineers, Adam Jacobson and Fan Zeng, are enjoying spending much of their time based at Caterpillar in the Brisbane CBD, becoming more experienced with the day-to-day operations of a major company in this space.
“We are currently preparing for a fourth major field trip, in which we will collect a larger amount of multi-sensor data than any of the previous trips, to be used for further development of both the tracking and the precise positioning systems.
“The innovative nature of the research has also been recognised by the research community: following up our paper published in the 2017 Australasian Conference on Robotics and Automation, we have just had a new research paper accepted at the 2018 IEEE/RSJ International Conference on Intelligent Robots, one of the premier international robotics conferences.
“Our research team will travel to Madrid, Spain to present the research in October this year. The paper: Adam Jacobson, Fan Zeng, David Smith, Nigel Boswell, Thierry Peynot, Michael J Milford, Semi-Supervised SLAM: Leveraging Low-Cost Sensors on Underground Autonomous Vehicles for Position Tracking, in IEEE/RSJ International Conference on Intelligent Robots, Madrid, 2018.
“The team’s work is also attracting regular media attention, including a recent feature piece in the Engineering and Technology publication on innovation in the autonomous mining sector in Brisbane. We have been invited to give a number of keynotes and presentations at mining and technology conferences across Australia: you will see us presenting at the Artificial Intelligence in Minerals & Petroleum Conference in Perth and Mine Automation in Brisbane. The project has also been highlighted in presentations to many of the top technology companies and universities around the world during overseas trips.
“We are looking forward to the remainder of the project, maturing the precise positioning systems, conducting increasingly gruelling field trials, and finalising the commercialisation hand-off to Caterpillar.”
