Max Robotics — Global Service Robot Directory + US Certification

About

Jiahui Zhu is a robotics researcher whose work bridges the gap between machine learning and real-world locomotion, focusing on how robots can learn to walk and move with the same developmental efficiency as humans. Her key research areas include quadruped and bipedal robot locomotion, Bayesian optimization, and developmental robotics—specifically, the concept of "scaffolded learning" inspired by human ontogeny. Zhu’s most notable contribution is pioneering the use of bootstrapping and scaffolding techniques to train bipedal walkers, where robots leverage simpler, supported tasks—like using training wheels or parental-like guidance—to progressively acquire complex walking skills, mirroring how human infants learn. Her 2019 paper on Bayesian optimization for quadruped robots (5 citations) demonstrated how to optimize 3D locomotion under real-world constraints, while her 2021 work on bootstrapping virtual bipedal walkers (3 citations) introduced a novel framework for achieving walking optimality by building on prior knowledge. Zhu also contributed to probabilistic inference methods for quadruped robots (2 citations), addressing the reality gap between simulation and physical hardware. Her research has been recognized for its innovative approach to reducing the challenges of applying machine learning to real-world robots, with a total of 12 citations across her most-cited papers.