Robot Operating System Anomaly Detection Module
Robot Operating System (ROS) has played a large part in advancing robotics research. It’s a versatile and popular middleware in the robotics community. Recently there have been growing security concerns, and a small number of works have offered preventive security mechanisms. We introduce a learning-based anomaly detection system for ROS, which can help augment the preventive methods with detection mechanisms. Specifically, we apply this concept of building a learning based model for industrial robotic arms that perform well-defined, repetitive tasks.
Our anomaly detection is called the ROS Anomaly Detection Module (ADM). This work goes over the details of the ADM and specific implementation details that not only assist with the ADM utilization, but also offers opportunities for building on existing features. In addition, this work also gives readers an idea of how the ADM can be adapted and employed alongside other projects.
Major Advisor: Rakesh Bobba
Committee: Jesse Walker
Committee: William Smart
Committee: Yeongjin Jang
Friday, August 31, 2018 at 1:00pm to 3:00pm
Kelley Engineering Center, 1007
110 SW Park Terrace, Corvallis, OR 97331