CV

Education/Certificates

• M.Sc. in Data Science, KTH, Stockholm, Sweden (09/21 - Now)
• B.Sc. in Mechatronics, Beijing Jiaotong University, Beijing, China (09/17 - 07/21)
• AWS certificated Data Analytics Specialty [Link to Badge]
• Microsoft Certified: Azure Data Fundamentals [Link to Badge]

Area of Speciality

• Robotics, Software Development, Data Engineering, Machine Learning, Computer Vision

Tools and Environment Protocols

• C/C++, Python, SQL, MATLAB
• Git, Docker, AWS …

Professional Experience

Automated Dice Pickup and Sorting System

Freelancer, UPWORK, Stockholm, Sweden

• The project aims to develop an automated system that uses a robot arm, a camera, and software tools to locate, pick up, and sort transparent red dice from a dice table.
• The dice will be picked up one at a time using a suction tool and placed into a box with 10 holes located on the table. The entire process will be guided by a webcam installed on the end of the robot arm, providing visual feedback.

Technology Stack: C++, Python, ROS, PyTorch, OpenCV, Git

Start-up Project Development

Developer, Stockholm, Sweden

• Collaborated with a Senior Data Engineer to develop a SQL generator utilizing LLMs like chatGPT and built an AWS data platform from scratch.
• Conducted performance evaluations of generated SQL queries and created a web application to seamlessly embed the SQL generation service within the platform.

Technology Stack: Python, AWS, SQL, Front-end development, Prompt Design for LLMs

Camera ISP optimization for computer vision tasks performed by DNNs

Master thesis intern, Scania, Södertälje, Sweden

• Working individually on the computer vision thesis project provided by Scania.
• Redefined the image segmentation task so the ISP in some way is part of the training process.
• Evaluated in what way the ISP affects the accuracy of the DNN output and if different ISP will influence the final accuracy.

Technology Stack: Python, PyTorch, OpenCV, Git, AWS

Dual-arm collaboration based on image recognition

Bachelor thesis student, Beijing Jiaotong University, Beijing, China

• Object recognition and 3D localization using a depth camera and YOLOv3. A visual-hand system is built with ROS.
• Once the object location and category have been determined, MoveIt will plan the movement and collaboration of the dual-arm robots’ arms. Robot arms will move following the planned trajectory to pick up cups and pour water.

Technology Stack: PyTorch, YOLOv3, ROS, MoveIt, Python, C/C++, TCP/IP

Analyze the data of the floor sweeper machine

Data Analyst Intern, Roborock, Beijing, China

• Roborock was developing a kind of cleaning robot that can be used in a shopping mall or underground parking lots.
• Responsible for visualizing and analyzing the data gathered from sensors and the battery, checking the data for any abnormalities, and writing a daily report.

Technology Stack: Python, Pandas, Matplotlib, ROS, Linux