Max Robotics — Global Service Robot Directory + US Certification

About

Xianglong Liu is a versatile researcher whose work spans robotics, autonomous systems, and multi-agent artificial intelligence. His research portfolio bridges two compelling domains: reliable robot navigation and the security of cooperative AI systems. In robotics, Liu made a notable early contribution with his work on visual loop closure detection for SLAM systems, developing a fast and incremental approach using proximity graphs that significantly improved upon traditional bag-of-words models in both speed and recall. This paper has accumulated 43 citations, establishing it as his most impactful work and a meaningful contribution to mobile robotics and autonomous navigation. More recently, Liu has turned his focus to the robustness and adversarial vulnerabilities of cooperative multi-agent reinforcement learning (MARL). His investigations into how adversarial minority agents can destabilize cooperative systems — and his subsequent work on mutual information regularization as a defense mechanism — address critical safety concerns before real-world deployment of multi-agent AI. He has also contributed to precision robotics hardware, exploring parametric modeling and singularity analysis of piezoelectric-driven parallel robots. Across these diverse threads, Liu's research consistently targets the reliability and resilience of intelligent autonomous systems, making his work increasingly relevant as robotics and AI move toward real-world deployment.