Shouwei Ruan

Papers

Total Citations

H-Index

About

Shouwei Ruan is a rising researcher at the forefront of embodied AI and adversarial machine learning, whose work critically examines the security vulnerabilities of next-generation intelligent systems. His most-cited paper (14 citations in 2024) pioneers the study of decision-level robustness in Large Language Model (LLM)-based embodied agents—systems that combine perception, language understanding, and physical action. Ruan’s key contribution lies in demonstrating how adversarial attacks can exploit the decision-making pipeline of these agents, revealing that even state-of-the-art LLMs, when integrated into robotic or virtual environments, are susceptible to subtle manipulations that alter their task planning and execution. This work bridges a critical gap between traditional adversarial robustness research and the emerging field of embodied AI, where safety and reliability are paramount. By systematically probing the intersection of language-driven reasoning and real-world interaction, Ruan has laid essential groundwork for developing more resilient autonomous systems. His research is particularly impactful for students and engineers building LLM-controlled robots, autonomous vehicles, or virtual assistants, as it underscores the need for robust, multi-level defenses in embodied architectures.