Xiaohang Liu

Currently, I am a Robotics Test Engineer with an M.S. in Artificial Intelligence and a B.S. in Computer Science. I am working toward a career in Robotics Software Engineering or Embodied AI, supported by a strong foundation in artificial intelligence, computer vision, machine learning, deep learning, and reinforcement learning, as well as the ability to quickly learn, adapt to, and apply new concepts and technologies.

Provide technical support and root-cause analysis for AMR and AGV deployments across North American project sites, troubleshooting issues spanning localization, navigation, perception, simulation, system integration, and hardware.
Design and develop AI-powered engineering tools that streamline troubleshooting, accelerate knowledge retrieval, and help field and engineering teams resolve technical issues more efficiently.

C++
Python
AMR / AGV
Localization
Navigation
Perception
Simulation
AI

Conducted comprehensive testing of Microsoft Azure Purview, validating data governance, security, compliance, and data lifecycle management capabilities across heterogeneous enterprise data environments.
Developed and automated SAP HANA test cases in C#, integrated them into CI/CD pipelines, and performed database validation across PostgreSQL, Oracle, MongoDB, Cassandra, and BigQuery.

C#
Big Data
SAP HANA
CI/CD
Azure Purview
Database Testing

Panda Vision-Based Color Sorting

- Built a ROS 2 Jazzy simulation where a Franka Emika Panda robot detects red, green, and blue boxes with an RGB camera, picks them with MoveIt 2, and sorts each box into its matching container. - Integrated Gazebo Sim, MoveIt 2, RViz, OpenCV, TF2, and a Python MoveIt interface in one workspace.

PackMind+ App (iOS & Android)

- PackMind+ is your ultimate companion for staying organized with personalized packing lists for any occasion. Create custom packing lists for travel, shopping, sports, and more with our intuitive interface. Never forget essential items again with reminders, visual tracking, and easy sharing capabilities.

Text-to-image generation for soccer tactics

- Developed a generative AI pipeline that transforms natural-language soccer tactics into realistic corner-kick scenes. Used SAM to segment and reposition players under tactical and spatial constraints, generated a synthetic image-caption dataset, and fine-tuned Stable Diffusion XL with LoRA to improve formation-aware image generation.

YUAttend Attendance App

- Developed an Android app using Java in Android Studio to track student attendance via GPS, displaying real-time location and attendance status for teachers. - Integrated Firebase for cloud data storage, allowing teachers to manage courses and export attendance records to Excel. - Planned for school-wide implementation, enabling all students to take attendance using the app.