iscover
Share this article on:
Powered by
MOMENTUM
MEDIA
An algorithm designed to prevent man-in-the-middle (MitM) attacks targeting unmanned military weapons and vehicles has been developed by Australian researchers.
The algorithm was developed as a joint initiative by AI professors at Charles Sturt University and the University of South Australia (UniSA), through an experiment that simulated human brain behaviour using deep learning neural networks to train a military robot’s operating system to detect the signature of a MitM attack.
A MitM attack refers to a type of cyber attack in which a threat actor positions itself between the initial commands and the receiving devices, allowing it to eavesdrop and collect information, as well as inject false commands.
“The robot operating system (ROS) is extremely susceptible to data breaches and electronic hijacking because it is so highly networked,” said UniSA Professor Anthony Finn.
“The advent of Industry 4, marked by the evolution in robotics, automation, and the internet of things, has demanded that robots work collaboratively, where sensors, actuators and controllers need to communicate and exchange information with one another via cloud services.
“The downside of this is that it makes them highly vulnerable to cyber attacks.”
Similarly, Dr Fendy Santoso of Charles Sturt Artificial Intelligence and Cyber Futures Institute says that the operating system used by unmanned vehicles, whilst having many benefits, ignores security issues.
The developed algorithm was trialled on a replica of US Army GVR-BOT combat vehicle and yielded successful prevention results of 99 per cent, with false positives occurring in under 2 per cent of cases.
The algorithm works by analysing the network traffic data of the robot, detecting attempts to intercept data transmission.
Finn said it works better than other world-recognised techniques for detecting cyber attacks.
“The speed of computing doubles every couple of years, and it is now possible to develop and implement sophisticated AI algorithms to guard systems against digital attacks,” he said.
Additionally, Santoso said that “owing to the benefits of deep learning, our intrusion detection framework is robust and highly accurate”.
“The system can handle large datasets suitable to safeguard large-scale and real-time data-driven systems such as ROS,” he said.
Santoso and Finn plan to continue testing the algorithm on a range of different unmanned vehicles with more complex operating systems and dynamics, such as drones.
“We are also interested in investigating the efficacy of our intrusion detection system on different robotic platforms, such as unmanned aerial vehicles, whose dynamics are reasonably faster and more complex compared to a ground robot,” said the IEEE report.
Be the first to hear the latest developments in the cyber industry.
