Enhancing Reliability in LLM-Integrated Robotic Systems: A Unified Approach to Security and Safety

2025-09-02

AI Summary

To address the insufficient reliability of LLM-empowered robotic systems under adversarial attacks and in complex environments, this paper proposes the first unified framework integrating safety (i.e., operational integrity) and security (i.e., resilience against malicious inputs). The framework incorporates dynamic prompt assembly, runtime state management, and multi-level safety verification mechanisms, alongside a dual-dimension evaluation metric—balancing performance and safety—that supports end-to-end validation in both simulation and real-world robotic platforms. Compared to baseline methods, our approach improves task success rate by 30.8% under prompt injection attacks and up to 325% in highly dynamic adversarial settings. These gains significantly enhance system robustness and deployability, thereby bridging a critical gap in the reliable integration of LLMs into embodied intelligence systems.

Technology Category

• Natural Language Processing: Safety and Robustness
• Intelligent Robots: Embodied AI
• Machine Learning: Adversarial Learning & Robustness

Application Category

• Search and Retrieval-Augmented AI: Search Tool Learning with LLM: Teaching LLMs to invoke search and make use of retrieved information
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

Integrating large language models (LLMs) into robotic systems has revolutionised embodied artificial intelligence, enabling advanced decision-making and adaptability. However, ensuring reliability, encompassing both security against adversarial attacks and safety in complex environments, remains a critical challenge. To address this, we propose a unified framework that mitigates prompt injection attacks while enforcing operational safety through robust validation mechanisms. Our approach combines prompt assembling, state management, and safety validation, evaluated using both performance and security metrics. Experiments show a 30.8% improvement under injection attacks and up to a 325% improvement in complex environment settings under adversarial conditions compared to baseline scenarios. This work bridges the gap between safety and security in LLM-based robotic systems, offering actionable insights for deploying reliable LLM-integrated mobile robots in real-world settings. The framework is open-sourced with simulation and physical deployment demos at https://llmeyesim.vercel.app/

Problem

• Enhancing reliability in LLM-integrated robotic systems
• Mitigating security risks from adversarial prompt injection attacks
• Ensuring operational safety in complex robotic environments

Innovation

• Unified framework mitigates prompt injection attacks
• Combines prompt assembling and state management
• Enforces safety through robust validation mechanisms