Sicheng Zhu

Papers

Total Citations

H-Index

About

Sicheng Zhu is a robotics researcher advancing intelligent assembly systems through deep reinforcement learning. His primary focus lies at the intersection of computer vision and robotic manipulation, particularly in solving the long-standing challenge of compliant peg-in-hole assembly—a critical process in precision manufacturing. Zhu’s most cited work, "A Visual Grasping Strategy for Improving Assembly Efficiency Based on Deep Reinforcement Learning" (2021, 5 citations), addresses a fundamental problem: the unpredictable fluctuations in contact force signals during attitude alignment, which degrade assembly accuracy. Rather than attempting to model these complex, uncertain disturbances, Zhu pioneered a visual grasping strategy that leverages reinforcement learning to adaptively correct alignment in real-time. This approach bypasses the need for precise force measurement, enabling more robust and efficient assembly operations. His research has direct implications for automated manufacturing, where reducing adjustment times and improving success rates are paramount. By integrating deep learning with traditional control methods, Zhu is helping to bridge the gap between simulation and real-world robotic dexterity. His work represents a meaningful step toward fully autonomous assembly systems capable of handling the variability inherent in industrial environments.