Security Considerations in AI-Robotics: A Survey of Current Methods, Challenges, and Opportunities

TL;DR

This survey addresses the security of AI-augmented robotics by presenting a three-dimensional taxonomy that covers attack surfaces, ethical/legal concerns, and Human-Robot Interaction security. It integrates analysis across perception, navigation/planning, and control layers, detailing physical and digital attack vectors, AI-model vulnerabilities, and defenses, while also examining roboethics, liability, privacy, and user trust. Key contributions include a comprehensive mapping of attack surfaces to robotic primitives, a synthesis of defenses across hardware, software, and AI components, and forward-looking directions in explainability, safe learning, and education. The work aims to guide researchers and practitioners toward robust, trustworthy AI-Robotics systems with improved resilience, accountability, and user acceptance in real-world deployments.

Abstract

Robotics and Artificial Intelligence (AI) have been inextricably intertwined since their inception. Today, AI-Robotics systems have become an integral part of our daily lives, from robotic vacuum cleaners to semi-autonomous cars. These systems are built upon three fundamental architectural elements: perception, navigation and planning, and control. However, while the integration of AI-Robotics systems has enhanced the quality our lives, it has also presented a serious problem - these systems are vulnerable to security attacks. The physical components, algorithms, and data that make up AI-Robotics systems can be exploited by malicious actors, potentially leading to dire consequences. Motivated by the need to address the security concerns in AI-Robotics systems, this paper presents a comprehensive survey and taxonomy across three dimensions: attack surfaces, ethical and legal concerns, and Human-Robot Interaction (HRI) security. Our goal is to provide users, developers and other stakeholders with a holistic understanding of these areas to enhance the overall AI-Robotics system security. We begin by surveying potential attack surfaces and provide mitigating defensive strategies. We then delve into ethical issues, such as dependency and psychological impact, as well as the legal concerns regarding accountability for these systems. Besides, emerging trends such as HRI are discussed, considering privacy, integrity, safety, trustworthiness, and explainability concerns. Finally, we present our vision for future research directions in this dynamic and promising field.

Figures (9)

• Figure 1: A generic robot architecture consisting of perception, navigation, and control stacks. Icons from flaticon
• Figure 2: An overview of our proposed taxonomy. The central parent node represents the overarching AI Robotic System, from which three primary branches emerge: Attack Surface, Ethical and Legal Concerns, and Human-Robot Interaction. Attack surfaces is further into two families: physical attack and digital attack. Physical attacks encompass attacks on the perception layer (input sensors). Digital attacks include attacks on navigation and planning, attacks on actuators, and attacks on AI models (training and inference attacks).
• Figure 3: An illustration of GPS spoofing attack on UAVs, in which an adversary redirects the UAV to a false location by transmitting deceptive GPS signals.
• Figure 4: An overview of attack surfaces in control systems and their impact on actuators within the operating environment. Icons from flaticon.
• Figure 5: An example of training attack on AI pipeline that causes misclassification of objects and misjudgment of object distance.