Junyu Nan

Papers

Total Citations

H-Index

About

Junyu Nan is a roboticist whose work bridges bio-inspired locomotion and reliable perception for autonomous systems. His research centers on two key areas: proprioceptive-inertial control for articulated robots and verification-driven deep learning for object detection. In his highly cited 2018 paper, "Proprioceptive-Inertial Autonomous Locomotion for Articulated Robots" (8 citations), Nan introduced a modular framework that leverages force sensing and inertial feedback—mimicking animals’ reliance on proprioception and vestibular cues—to enable adaptive gait. A standout contribution is the concept of "anti-compliance," a novel application of positive force feedback that allows robots to navigate uncertain terrain without external sensors. Complementing this, his 2019 work "Combining Deep Learning and Verification for Precise Object Instance Detection" (2 citations) tackles the critical problem of false positives in deep learning detectors. By integrating formal verification with high-confidence detections, Nan proposes a system that prioritizes reliability over generic mAP metrics—a vital step toward deploying robots in safety-critical environments. Though early in his career, Nan’s work demonstrates a rare synthesis of control theory, biomechanics, and trustworthy AI, positioning him as a promising voice in autonomous systems research.