Max Robotics — Global Service Robot Directory + US Certification

About

Wenjia Niu is a leading researcher at the intersection of artificial intelligence security and reinforcement learning, with a particular focus on adversarial robustness in autonomous systems. Her pioneering work investigates how deep reinforcement learning (DRL) systems—critical for applications from robot control to natural language processing—can be vulnerable to carefully crafted adversarial attacks. In her highly cited 2018 paper on adversarial examples in DQN pathfinding training (25 citations), Niu demonstrated how white-box Q-table variations can be exploited to mislead autonomous navigation systems. She further advanced this line of inquiry with a PCA-based predictive model for adversarial examples in Q-learning pathfinding (23 citations), providing a systematic framework for understanding attack vectors. More recently, Niu has developed curricular robust reinforcement learning approaches using GAN-based perturbations (2022, 12 citations), introducing continuously scheduled task sequences to train more resilient autonomous systems. Her work has direct implications for the safety and reliability of cooperative robots, autonomous vehicles, and distributed AI systems. By systematically identifying vulnerabilities in path planning and proposing novel defense mechanisms, Niu is helping to establish the foundational principles for trustworthy AI in safety-critical applications.