CV

Algorithm Engineer

Xingyun Intelligence (Shenzhen) Technology Co., Ltd.

Key contributor to the development and productization of autonomous driving systems — Autonomous Forklift, AGV, and IGV — from prototype to real-world deployment in ports and factory environments.

• Built core perception, SLAM, planning, and control modules with multi-sensor fusion pipelines (camera, LiDAR, GNSS/RTK, IMU) on ROS/C++, deployed on production fleets.
• Collaborated cross-functionally across hardware, software, and operations teams; contributed to project management of multi-vehicle deployments.

M.Sc.

University of Gothenburg

Computer Science and Engineering

• Thesis: Assessment & Failure Recovery in Remote Vision-Language-Action (VLA) Deployment
• Supervisor: Ze Zhang | Examiner: Ahmed Ali-Eldin Hassan | Thesis partner: Yajing Zhang
• Focus: Deep Learning and Machine Learning
• Selected courses: Deep Machine Learning, Advanced Topics in Machine Learning, Design of AI Systems, Computational Semantics, Discrete Optimization, Advanced Algorithms

B.Eng.

Guangzhou University of Chinese Medicine

Medical Information Engineering

• Thesis: The Application of Ensemble Learning in Quantitative Investment Prediction
• Outstanding Graduate Award, 2017
• Core courses: Data Structures & Algorithms, Java, C, Database Design, Networking, Operating Systems

Simultaneous Indoor Tracking and Activity Recognition Using Pyroelectric Infrared Sensors

Sensors (MDPI), 17(8):1738

3rd author. X. Luo, Q. Guan, H. Tan, L. Gao, Z. Wang, and X. Luo. Proposed a method for simultaneous indoor tracking and activity recognition using PIR sensor arrays and a two-layer Random Forest classifier.

Abnormal Activity Detection Using Pyroelectric Infrared Sensors

Sensors (MDPI), 16(6):822

2nd author. X. Luo, H. Tan, Q. Guan, T. Liu, H. Zhuo, and B. Shen. Proposed an unsupervised method for abnormal activity detection using spectral clustering, Hidden Markov Models (HMMs), and One-Class SVMs.

Intelligent Guided Vehicle (IGV) — Autonomous Port Truck (ART)

Developed multi-LiDAR calibration and multi-sensor fusion for LiDAR-based obstacle avoidance and localization, achieving a throughput of 35 containers/hour in port operation.

• Multi-LiDAR calibration and sensor fusion (LiDAR, GNSS/RTK, IMU)
• Achieved 35 containers/hour throughput in real-world port deployment
• Video: youtube.com/@yunxing-port/videos

Autonomous Forklift

Implemented LiDAR–camera calibration and MPC-based motion planning for accurate automated loading and unloading operations.

• LiDAR–camera calibration and sensor fusion
• MPC-based motion planning for precise navigation
• Achieved 90–92% success rates in automated loading and unloading

Automated Guided Vehicle (AGV)

Built a LiDAR–IMU SLAM system with loop closure for mapping, localization, and map extension in indoor and outdoor factory environments.

• LiDAR–IMU graph-based SLAM with loop closure
• Deployed across indoor and outdoor factory environments

Indoor Activity Recognition with Pyroelectric Infrared Sensors

Research on simultaneous indoor tracking and abnormal activity detection for elderly assisted living, resulting in two publications in Sensors (MDPI).

• Pyroelectric Infrared (PIR) sensor array with FOV modulation
• Two-layer Random Forest classifier for tracking and activity recognition
• Abnormal activity detection via HMMs and One-Class SVMs
• Published in Sensors (MDPI), 2016 & 2017