Dr Branko Ristic – investigating possibilities in data fusion Thursday, 08 October 2015

Dr Branko Ristic (Department of Defence), Senior Research Fellow, RMIT University, School of Electrical & Computer Engineering

Sponsored editorial piece written by RMIT University, Principal Partner of Engineers Australia Victorian Division.

Dr Branko Ristic from the Department of Defence has joined RMIT University as a Senior Research Fellow in the School of Electrical and Computer Engineering. He is an accomplished researcher with over 30 years experience in universities, government laboratories and industry. His appointment was made possible by the Defence Science and Technology Group’s Industry Placement Scheme.

What is your area of expertise?
My area of research speciality is data fusion (also known as Information fusion), in particular tracking and multi-sensor data fusion.

What is data fusion?
Data fusion is an indirect process of making inference (and drawing conclusions) from observational data. It starts with a system that observes its environment. This could use surveillance sensors, such as radar, sonar, video, or radiological or chemical sensors. We then apply mathematical models to the data of how it relates to the subject we are interested in to extract new useful information. A critical aspect of data fusion is combining data from different sensing technologies to improve the robustness, specificity, scope and/or accuracy of the result.

How has it been used in industry?
Data fusion has a very broad range of applications. Most systems capable of sensing and computation have the potential to use data fusion techniques.

Examples of how data fusion is used include tracking underwater targets (e.g. submarines) using sonar systems; tracking people in crowded areas using surveillance cameras (railway stations, airport terminals); localisation of sources of dangerous gas emissions in urban environment, detection and tracking of anti-ship missiles using infra-red sensors, to name a few.

Sensor fusion is also related to machine learning, where historical data are used to learn patterns of behaviour. This can be used to detect anomalies, for example in modelling the behaviour of traffic in commercial ports by combining radar and Automatic Identification System (AIS) data.

What research will you be undertaking at RMIT?
Initially, I am undertaking two research projects at RMIT. The first investigates the use of multistatic sonar buoys to track submarines. The second is to develop search techniques for autonomous systems investigating chemical, biological or radiological accidents.

What are the trends and challenges in your area of research?
New sensing technologies and applications are under constant development. The last decade saw huge progress in biological sciences applications of data fusion (e.g. epidemics prediction, DNA sequencing). I see the next big application of data fusion being autonomous systems. It will be a key component in helping these systems make decisions about how to act.

RMIT University