Empirical Assessment of Methods to Detect Cyber Attacks on a Robot

摘要

An experiment was conducted using a robot to investigate the effectiveness of four methods for detecting cyber attacks and analyzing robot failures. Cyber attacks were implemented on three robots of the same make and model through their wireless control mechanisms. Analysis of experimental data indicates the differences in attack detection effectiveness across the detection methods. A method that compares sensors values at each time step to the average historical values, was the most effective. Further, the attack detection effectiveness was the same or lower in actual robots as compared to simulation. Factors such as attack size and timing, influenced attack detection effectiveness.