School of Mechanical & Mining Engineering

The Smart Machines Group undertakes research into  robotic and automonous machinery with a particular emphasis on  control systems, mission planning, environment perception, and human machine interaction.  We are the home of the Cooperative Research Centre for Mining (CRCMining) Automation Program and much of what we do targets the realization of autonomous mining systems such as robotic excavators, trucks, and bulldozers. Our work is largely focussed on addressing four gaps that limit the ability to deploy sophisticated robotic systems in real world applications:
  • Control strategies that enable automated machines to operate interdependently with other equipment, both manned and automated, in semi-structured environments;
  • Optimization algorithms that find the most efficient way to break a high level task description into a detailed action plan;
  • Situational awareness capabilities that are able to replace the many and varied functions performed by human operators; and
  • Perception systems that enable the effective remote operation of machines over long distances where communication bandwidths are limited and where there may be significant latency.
These are multi-disciplinary engineering challenges encompassing perception, control, planning, state and parameter estimation, software engineering, reliability engineering and human factors.
 
The group currently has about twenty-five members and is based in the School of Mechanical and Mining Engineering at the University of Queensland. We also operate a laboratory near Caboolture where we conduct research with full-scale surface mining machines.
 
The Smart Machines Group conducts research in:
  • Model Predictive Control for large machinery
  • Mission planning  using receeding optimization methods
  • Perception and control requirements for teleoperation of machinery
  • Situational awareness for automation
  • The applicatoin of functional safety in automation systems

Associated Research Centres

CRC Mining

Staff

Program Leader
Professor Ross McAree

Dr Michael Kearney

Dr Kevin Austin

Dr Zane Smith

Contact Details

Please direct all enquiries here.

Current Research

Mining Shovel Semi Automation This aims to provide operator assists that improve truck-shovel operations. The benefits of the project span safety, availability, productivity, and maintenance.  The project is developing, inter alia, a layer-of-protection capability that prevents metal-on-metal collision that might injure truck drivers and damage trucks and an automated loading capability that reduces the cycle time, lowers machine duty through smoother operation, and reduces the workload on operators.
 

Minimal perception requirements to support effective remote control of bulldozers  This project aims to understand and define operator perception requirements for effective remote operation of bulldozers on coal stockpiles and determine how the feedback information to meet these needs can be optimized to fit within the capacity of contemporary wireless communication channels.  
 

Real-time receding horizon mission planning for automated excavation The aim of this research is to develop algorithms and techniques for generating optimal excavation plans for diggers. This is seen as a key step towards the generation of mission plans for automated diggers.      

Perception sensor capabilities This project aims to establish a formal framework for evaluating sensors against requirements relevant to typical mining automation applications. A selection of candidate sensors is being evaluated, including: an Indurad radar; Velodyne HDL 64E high-definition LIDAR, SICK LD-MRS, SICK LRS3100 and SICK LMS511 scanning LIDARs and XIMEA RL13C cameras.

Mining Automation Reference Architecture The motivation for this project is the compelling need for of a common plan for bringing disparate mining technology together to form a compatible and effective mine-wide automation solution. We call this common plan the Mining Automation Reference Architecture (MARA) and it broadly seeks to capture the essence of existing architectures, and the vision of future needs and evolution to provide guidance to assist in developing new system architectures.