Centre for Intelligent Systems
The Centre for Intelligent Systems (CIS) conducts outcome driven as well as theoretical research in the areas of complex intelligent systems. A particular focus is applied technologies relating to ICT and Engineering applications.
The Centre’s major research concentration is around the strong research clusters from Information and Computing Sciences and Engineering. The core research areas within CIS are:
- Computational Intelligence
- Simulation, Automation and Robotics
- Smart Networked Devices
- Clean Energy Technologies.
The Centre is currently involved in a number of different research projects, one of these is the Automatic System for Roadside Fire Risk Assessment. The aim of this project is to develop an automatic system that can analyse roadside video data and assess roadside fire risks. A number of novel computational intelligence and pattern recognition based techniques for identification of roadside objects from video frames and assessing fire risks are developed. Firstly, the segmentation and classification techniques to extract brown grass from roadside video frames are developed. Secondly, the techniques to calculate the coverage and length of grass are developed. Finally, fire risk assessment techniques to calculate biomass and to identify fire risks using grass coverage and grass length are developed. View more information on roadside video data collected using an automated vehicle.
The main problem is that there is no system for fire risk assessments, so not only here in Australia but nowhere, there's no automated system. Currently they're using manual systems which are time-consuming, costly and also they do not use DVR, that digital video recording the data which Transport and Main Roads they record every year, all Queensland roads. So we are trying to develop a novel automated system which can tell about fire risk that low, medium, as I said low, medium and high fire risk.
A success for us means if we can get an automated system which can produce accuracies as good as or better than human observations, because we were working on this project for a couple of years, so we have developed techniques for segmentation and classification and we had investigated those techniques. Because those are very important, because they are used in calculating those areas, grass areas and height of the grass, so that we can tell whether for high fire risk or low fire risk. So we have also compared those techniques with other techniques. So we have been doing that and we have got good techniques now, which we applied in our system.
Every landowner has the responsibility to look after the fire loads and fire risks on their land and many fires start from roads due to cigarette butts, or vehicles, or other reasons and so Main Roads has a program to monitor and manage near the fire risk in their road corridors. Well since I approached Brijesh to work with Main Roads about eight years ago, we’ve put together a number of grants and contracts with CQU to develop this artificial intelligence approach to analysing video for fire risk.
So first our project with Transport and Main Roads was like a contract, they contracted me to physically go to all Central Queensland roads to collect data, so when we visited all Queensland, all Central Queensland roads so we collected data. When I say data, means not video data, physical data. We chopped grass within one square metre, we brought to Uni and we calculated biomass and physically because we needed that to compare our techniques and also we did a survey, like humans, so three experts. We were driving through and for example, saying okay here is high grass, it's low grass, here is a high biomass and also we did a whole survey. So that started in 2013 and when we collected that data, then we wanted to develop an automated system after that. We would like to improve our techniques further. We focussed only on vegetation, but we want to extend those techniques so that they can detect other objects also, which can be later applied to other safety applications, so that that's what our next step is.
CQU have developed a methodology for assessing accurately the fuel loads on the roads, that is starting to be applied in regions of the state so that we can get an annual assessment of the fire risk very cheaply and on a timely basis and then the methods are also being applied to other areas such as road safety where they have even greater potential benefits. CQU represents possibly the only local group that would be able to do this type of artificial intelligence development work with Main Roads. Really they’re a critical partner to making this happen.
Get Involved in our Research
CIS is committed to outcome driven research and its applied research is well connected to local industry such as Transport, Energy, Health and Mining. CIS is well recognised in its focused research areas and its strength is in the following areas.
Artificial/Computational Intelligence, Pattern Recognition and Data Mining, Image Processing, Computer Vision, Applied Maths.
Simulation, Automation & Robotics
Intelligent Modelling, Intelligent Robotics, Intelligent Sensing Decision Support, Mining and Medical Applications.
Smart Networked Devices
Smart Homes, Smart Grids, Intelligent Protection Systems, Network Automation.
Clean Energy Technologies
Intelligent Buildings, Alternate Fuel, Sustainable HVAC, Materials and Processes, Flow of Energy Modelling.
RESEARCH HIGHER DEGREE STUDENTS
CIS is proud to host and provide research higher degree training for postgraduate students. CIS has a number of research higher degree students based in Rockhampton, Brisbane and Sydney campuses. CIS provides excellent multidisciplinary research training environment.
Scholarships are called twice per year for the general CQUniversity merit rounds. In addition, projects are approved from time to time which allow stipend support for students to study on specific Centre relevant research problems. Further information is available on our Research Scholarships page.
CIS has expertise in:
- Artificial/Computational Intelligence
- Pattern Recognition and Data Mining
- Image Processing
- Computer Vision
- Applied Maths
Simulation, Automation & Robotics
- Intelligent Modelling
- Intelligent Robotics
- Intelligent Sensing
- Decision Support
- Mining and Medical Applications
Smart Networked Devices
- Smart Homes
- Smart Grids
- Intelligent Protection Systems
- Network Automation
Clean Energy Technologies
- Intelligent Buildings
- Alternate Fuel
- Sustainable HVAC
- Materials and Processes
- Flow of Energy Modelling
Our research projects include:
- ARC - Smart information processing for roadside fire risk assessment
- ARC - Improve fault detection in electrical distribution networks to avoid catastrophic bush fires
- ARC - Networked and coordinated control methods
- ARC - Framework for optimised ensemble classifiers
- Advance Queensland - Battery and microgrid management systems
- ARRB – Impact of wide centre lines on the pavement using image processing techniques
- Digital mammography: classification of benign and malignant patterns
- Opening access to valuable documents: a handwriting recognition approach
- Elevare energy dstatcom project
- Hanging wire detection
- High performance swer earthing
- Power electronics projects in power quality and PV
- Failure analysis of CV 103 pulley and bearing for BMA
- Thermo-chemical conversion of non-recyclable waste plastics into fuel.
- Development of a novel low energy building using bio phase change materials and rooftop greenery system.
- Technology development for sustainable use of storm water for irrigation purposes in agricultural farm land.
- Energy recovery and power generation from solid wastes (MSW, Green waste, wood waste, sugarcane bagasse, etc.)
- Modelling and Analysis of the Processes of Dead-Burned Magnesia Production in Order to Improve Energy Efficiency and Environmental Sustainability
- Development of non-adiabatic Desiccant Wheel for Solar Thermal Air Conditioning
- Technology development for solar assisted desalination
- Integration of industrial processes in an existing coal fired power station to minimize the cost of carbon capture and storage
- Flow phenomena in the transportation of mineral slurry in open channel launders.
- Team programming: framework development
- Team Programming: Application Development
- Gamified training materials for drug prescribing
The Centre for Intelligent Systems undertakes both basic and applied research, with a particular focus on collaborative partnerships to deliver real world solutions.
CIS partners include:
- Atlass Pty Ltd
- Elevare Energy
- Intendico Pty Ltd
- Queensland Health
- Queensland Transport and Main Roads
- Western Power