Alvin Zhu
B.S. Computer Engineering,
University of California, Los Angeles
Research Assistant at UCLA Robotics and Mechanisms Laboratory (RoMeLa)
I have published 6 research papers, including 2 first author papers at ICRA and 1 first author paper at Humanoids, with another first-author paper (presented at CoRL 2025 Workshops) currently under submission to ICRA. Over the past two years, I have also played key roles in multiple cutting-edge projects as part of a graduate robotics research lab under the guidance of Prof. Dennis Hong. I continued my robotics journey this past summer as a robotics platform software intern at NVIDIA, where I worked on applying vision language models (VLMs) for multi-agent robot collaboration. My research and projects lie at the intersection of robotics hardware, robot learning, simulation, and robotic perception, with a focus on developing intelligent and adaptive robotic systems. I am passionate about leveraging deep reinforcement learning and LLM/VLM agents to optimize control policies for platforms like legged robots and manipulators, and bridging the sim-to-real gap for zero-shot deployment.
Selected Publications
AURA: Autonomous Upskilling with Retrieval-Augmented Agents
Alvin Zhu*, Yusuke Tanaka*, Andrew Goldberg, and Dennis Hong
Accepted to 2025 Conference on Robot Learning (CoRL) Workshops and in review for 2026 IEEE International Conference on Robotics and Automation (ICRA)
​
​
​
​
Mechanical Intelligence-Aware Curriculum Reinforcement Learning for Humanoids with Parallel Actuation
Alvin Zhu*, Yusuke Tanaka*, Quanyou Wang, and Dennis Hong
Accepted to 2025 IEEE-RAS International Conference on Humanoid Robots (Humanoids)
​
​
​
​
ICRA
Cycloidal Quasi-Direct Drive Actuator Designs with Learning-based Torque Estimation for Legged Robotics
Alvin Zhu*, Yusuke Tanaka*, Fadi Rafeedi, and Dennis Hong
Accepted to 2025 IEEE International Conference on Robotics and Automation (ICRA)
​
​
​
​
ICRA
AeroConf
Mechanisms and Computational Design of Multi-Modal End-Effector with Force Sensing using Gated Networks
Alvin Zhu*, Yusuke Tanaka*, Richard Lin, Ankur Mehta, and Dennis Hong
Accepted to 2025 IEEE International Conference on Robotics and Automation (ICRA)
​
​
​
​
Tethered Variable Inertial Attitude Control Mechanisms through a Modular Jumping Limbed Robot
Yusuke Tanaka, Alvin Zhu, Dennis Hong
Accepted to 2025 IEEE Aerospace Conference (AeroConf)
​​
​
​
​
Robotics Projects Overview
Humanoid Locomotion using Deep RL and AI Agents
I created GPU-accelerated simulation and training pipeline for the humanoid robot BRUCE using MuJoCo-MJX, enabling efficient deep reinforcement learning for dynamic and stable locomotion. The system leverages parallel environments and accurate modeling of complex mechanisms to train policies that remain robust under external disturbances. To automate the training process, I integrate AI agents and retrieval-augmented generation (RAG) into a multi-agent framework that iteratively constructs and refines the curriculum. This approach moves toward fully prompt-to-policy humanoid training, reducing human effort while improving generalization and zero-shot transfer to real-world deployment.
Advanced Robotic Perception System for Humanoid Soccer
I developed the full perception stack for the humanoid robot ARTEMIS, enabling full spatial awareness in dynamic RoboCup soccer environments. It integrates the YOLOv8 deep learning model and classical computer vision algorithms with point clouds for object detection, 3D pose estimation, and proximal object detection robust to heavy amounts of noise, allowing ARTEMIS to aim and score 45 goals in 6 seated matches and overthrow the reigning champions 6 goals to 1.
Object Segmentation using Vision Transformers and Deep Learning models
I integrated the Segment Anything Model (SAM) vision transformer with custom YOLOv8 detection weights to provide 95% accurate segmentation of slide handles and stairs. The segmented object's positions are extracted from the Intel Realsense D435 camera's point cloud for use in simultaneous locomotion and grasping.
Cost Efficient 3D Printed Robot Dog
The robot dog project involved designing and developing a fully functional quadruped robot. I 3D modeled and manufactured the upgraded big dog, ensuring a compact and efficient mechanical design optimized for strength and cost efficiency. I implemented a PID control system integrated with an IMU to enable real-time balance and stability.