State of the Art Tech Improves Intruder Tracking Surveillance

This article is a repost from DSI.



For 24-hour surveillance to be effective for Australia’s national security there can be no room for human error.


To eliminate this human element from critical but often tedious surveillance, robotics specialists Agent Oriented Software (AOS) sought specialist research expertise and teamed up with Queensland University of Technology (QUT). With a boost from a Defence Science Institute (DSI) Collaborative Research Grant, the team developed its current prototype of an autonomous intelligent intruder tracking system named ‘eyeSight’.


Through Artificial Intelligence (AI) and robotics, eyeSight can detect an object in a panorama video, identify whether the object (person or vehicle) has authorisation to enter the premises and track the GPS location of the object using footage from multiple cameras. Combined with an unmanned ground vehicle (UGV), it can safely approach and inspect an intruder, enhancing military base security.


Originally envisaged as a collaboration between AOS and the Australian Research Council (ARC) Centre of Excellence for Robotic Vision headquartered at QUT, the project evolved into a partnership with the university’s School of Electrical Engineering and Robotics. The team was made up of software engineers from AOS, led by Managing Director Dr Andrew Lucas, and researchers from QUT including select early-career post-doctoral students, who gained valuable industry experience.


“The grant from DSI expanded AOS’s innovation capability and encouraged us to experiment with new ideas. DSI’s deep network of relationships within the Defence sector, many top-tier universities and local companies helped us find the right research partner in QUT,” said Weixun Leon Liu, a Software Engineer at AOS who worked extensively on the project.


Dr Simon Denman, Computer Vision Specialist and Senior Lecturer at QUT’s School of Electrical Engineering and Robotics, oversaw the computer vision call aspect of the project. “As someone who works within a university, it was nice to be able to apply research translation towards a project with real-world use,” he said.


“Universities are pushing the boundary of technological advancement. Engagement helps industry to better plan and adjust their development directions,” said Weixun.


eyeSight will greatly improve the autonomous tracking capability for AOS’s unmanned ground vehicles, currently being developed as part of the ADF and Australian Army’s Jericho Dawn program - a vision to develop a future force that is agile and adaptive.


eyeSight's story of research progression is one of many at DSI and demonstrates how research outcomes are transferable and of value across multiple sectors, including defence. DSI aims to establish connections between industry and academia and encourages researchers who may not have had any experience with the defence sector to make contact to leverage the many grant and support options available to advance their research.


To learn more about this project, or any other research collaboration projects facilitated by DSI, please contact us at dsi.info@defencescienceinstitute.com.