Professors from the University of South Australia and Charles Sturt University have developed an AI algorithm to detect and counter man-in-the-middle (MitM) attacks on autonomous military robots.
See also: How is artificial intelligence (AI) affecting the job search?
MitM attacks are a type of cyberattackwhere the data traffic between two parties, in this case the robot and its legitimate controllers, is monitored either for espionage or to insert fake data into the stream. Such malicious attacks aim to disrupt the operation of autonomous vehicles, modify the transmitted instructions and in some cases, even take control, teaching the robots to perform dangerous actions.
“The robot operating system (ROS) is extremely vulnerable to data breaches and cyber-attacks because it is so highly networked,” commented Professor Anthony Finn, who participated in the study.
“The advent of Industry 4, marked by the evolution of robotics, automation and the Internet of Things, has required robots to work collaboratively, with sensors, actuators and controllers needing to communicate and exchange information with each other via cloud.“
“The downside to this is that it makes them highly vulnerable to cyberattacks.“
Researchers at the university have developed an algorithm that uses machine learning techniques to detect and counter these attempts with extreme speed. The algorithm’s effectiveness has been proven through rigorous testing by the United States. The military’s GVR-BOT (TARDEC) has had impressive success, with a 99% in preventing attacks and false positives of less than 2% in tested scenarios.
See also: Google: AI image creation directly in the Search bar
MitM detection in unmanned vehicles and robots is a complex process, as these systems operate in fault- tolerant ways . Thus, the distinction between normal operation and fault conditions can be unclear. Furthermore, robotic systems can be compromised at various levels, from the core of the system to its subsystems and sub-components, causing operational problems that can disrupt the robot's operation.
Researchers at the university have developed a system that analyzes robot movement data to detect tampering attempts. This system uses node-based methods, examines data , and uses a flow statistics-based system that reads metadata from the packet header.
“We are also interested in investigating the effectiveness of the intrusion detection system on various robotic platforms, such as those of autonomous aircraft, whose dynamics are relatively faster and more complex compared to a ground robot,” concludes the research published on the IEEE website.
What are the main advantages of using Artificial Intelligence algorithms to detect MitM attacks?
MitM attacks remain a significant threat to the security of autonomous military vehicles. The application of Artificial Intelligence algorithms to detect these attacks offers a number of key advantages.
See also: Microsoft: New bug bounty program for AI-powered Bing
Flexibility and Adaptability
First, AI algorithms can adapt and learn from new data, allowing them to develop an ever-evolving understanding of threats. This allows them to react to changing attack tactics and new forms of threats.
Automation and Speed
Second, the automation offered by AI algorithms enables rapid detection and response to threats. Rapid response can prevent further damage and quickly restore system security.
Specialization
Third, AI algorithms can be specialized for specific applications, such as autonomous military vehicles. This creates an environment where the specific requirements and risks of these systems have been taken into account.
Source: bleepingcomputer
