Xiaopeng (Rock) Wu

I am a graduate researcher in Computer Science at Binghamton University, working in the Autonomous Intelligent Robotics (AIR) Lab under the supervision of Prof. Shiqi Zhang. My research focuses on embodied AI and humanoid robotics.

I work on vision-based whole-body teleoperation, embodied data collection, Vision-Language-Action model training and deployment, sim-to-real evaluation, and autonomous behaviors for humanoid robots.

Before graduate research, I spent eight years building production software and large-scale data platforms, including four years at PayPal. I bring software engineering, distributed systems, and production reliability experience to robotics research.

Goal: Build reliable humanoid systems that learn from high-quality real-world interaction data.

Research Focus: Humanoid teleoperation, embodied data, VLA systems, sim-to-real evaluation, and robot learning infrastructure.

Current Project: G1 Teleop Data Platform.

News

Current Work

2026

G1 Teleop Data Platform

A modular, vision-based platform for Unitree G1 teleoperation, embodied data collection, and VLA training and deployment. The system connects human demonstrations, synchronized robot data, policy development, and on-robot evaluation.

Private repository; public documentation and results are in preparation.

Research Interests

Humanoid robotics, embodied AI, robot teleoperation, multimodal robot data, Vision-Language-Action models, sim-to-real transfer, robot learning infrastructure, autonomous navigation, and systematic evaluation on physical robots.

Experience

Graduate Student Researcher

Autonomous Intelligent Robotics (AIR) Lab, Binghamton University ยท Advisor: Prof. Shiqi Zhang

Research in embodied AI and humanoid robotics, including whole-body teleoperation, data collection, VLA systems, sim-to-real experiments, autonomous behaviors, and physical robot validation.

Software Engineer

PayPal, Shanghai

  • Built a Java, Spark, and Hadoop platform processing more than 200 TB of data daily.
  • Led a five-person team migrating 2 PB of enterprise data to Google Cloud Platform.
  • Developed resource-allocation tooling that reduced distributed job failures by 60%.
  • Maintained more than 99% data consistency during cross-platform migration.

Software Engineer

PingAn Medical & Healthcare, Shanghai

  • Built a data-lineage-driven ETL platform that replaced more than 500 scripts.
  • Designed a distributed scheduler supporting up to 1,500 peak tasks.
  • Implemented low-latency CDC pipelines handling more than 1 TB daily.