Australian researchers have harnessed artificial intelligence to identify and thwart man-in-the-middle (MITM) attacks on the widely-used open-source Robot Operating System (ROS).

As detailed in an IEEE publication, the researchers executed a penetration test on the American military robot, GVR-BOT, which operates ROS on its onboard computer, to pinpoint vulnerabilities, ItNews reported.

During the test, they collected network data from the robot operating in both “legitimate and malicious conditions.” This data was then utilized to train a convolutional neural network (CNN) to distinguish attack traffic.

Researchers from the University of South Australia (UniSA) and Dr. Fendy Santoso from the Charles Sturt University AI and Cyber Futures Institute collaborated with the US Army Futures Command to recreate the MITM attack.

The results were promising, with the CNN algorithm, operating on ROS, achieving a remarkable 99% success rate in preventing malicious attacks, as reported in a statement by UniSA.

Written by staff