Fingerprinting Movements of Industrial Robots for Replay Attack Detection
Hongyi Pu, Liang He, Chengcheng Zhao, David K. Y. Yau, Peng Cheng, Jiming Chen
Abstract
Industrial robots are prototypical cyber-physical systems widely deployed in (smart) manufacturing, which operate according to the operation code uploaded by the human operator and are monitored in real-time based on their movement data. However, industrial robots suffer from replay attacks, via which attackers can manipulate the robot operation without being observed by the monitoring system. To mitigate this vulnerability, we design a novel intrusion detection system for industrial robots using their power fingerprint, called <monospace xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">PIDS</monospace> ( <underline xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">P</u> ower-based <underline xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">I</u> ntrusion <underline xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">D</u> etection <underline xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">S</u> ystem), and deliver <monospace xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">PIDS</monospace> as a <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">bump-in-the-wire</i> module installed at the powerline of commodity robots. The foundation of <monospace xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">PIDS</monospace> is the physically-induced dependency between the robot movement and the concomitant power consumption, which <monospace xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">PIDS</monospace> captures via joint physical analysis and (cyber) data-driven modeling. <monospace xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">PIDS</monospace> then fingerprints the robot movements observed by the monitoring system using their expected power consumption, and cross-validates the fingerprints with empirically collected power information — a mismatch thereof flags anomalies of the observed movements (i.e., evidence of replay attack). We have evaluated <monospace xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">PIDS</monospace> using three models of robots from different vendors — i.e., ABB IRB120, KUKA KR6 R700, and Universal Robots UR5 robots — with over 2,000 operation cycles. Experimental results show that <monospace xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">PIDS</monospace> detects replay attacks at an average rate of 96.5 percent (up to 99.9 percent) and a 0.1s latency.