Xuanyu Fang

Cornell University, Shanghai Normal University

Papers

Total Citations

H-Index

About

Xuanyu Fang is a robotics researcher whose work bridges human-robot interaction and autonomous navigation. Her key research areas include failure detection in social robotics, multi-sensor fusion for localization, and set-membership estimation. Fang’s major contribution is the creation of the Bystander Affect Detection (BAD) Dataset, a novel resource that enables robots to detect their own errors by reading the implicit social cues—confusion, smirks, or giggles—from human bystanders. This work, published in 2023, has already garnered 5 citations and represents a paradigm shift in how robots can self-correct without explicit feedback. In 2025, Fang advanced mobile robotics with a zonotopic set-membership estimation approach that fuses GNSS, odometry, and 2D laser data for seamless indoor-outdoor localization, earning 4 citations. Her method ensures robust, bounded-error positioning even in challenging transitional environments. Fang’s research is notable for its practical impact: the BAD Dataset opens new avenues for socially aware robots, while her localization work directly supports real-world deployment across varied terrains. Her interdisciplinary approach—combining affective computing with rigorous estimation theory—positions her as an emerging leader in creating robots that are both perceptive and reliable.