About Me | KaguraTart

👤 About

Ph.D. candidate at Beihang University (BUAA), School of Transportation Science and Engineering. M.Sc. at Tongji University, College of Transportation Engineering. Previously interned at Huawei Vehicle BU and Bosch China.

🎓 Education

Beihang University (BUAA)2024 – Present

Ph.D. · Transportation Science & Engineering

Tongji University2021 – 2024

M.Sc. · Transportation Engineering

Courses: Computer Aided Design, Transportation Data Analysis, Numerical Computation, Data Mining

Tongji University2017 – 2021

B.Eng. · Civil Engineering

Courses: Finite Element Method, Calculus, Linear Algebra, Probability & Statistics, C++

Awards: 2020 Campus Scholarship, 2018 Kaisa Scholarship

💼 Experience

Shanghai Yinwang Intelligence (Huawei Vehicle BU)2024.07 – 2025.08

DevOps Engineer

Extract CAN bus signals from onboard .MF4 data, parse and generate core dashboard JSON files
Visualize vehicle perception data, convert to OpenX format and import into CarMaker for ADAS SIL testing
Design automated test cases for autonomous driving and parking edge cases using CarMaker and CARLA

Bosch (China) Investment Ltd.2023.07 – 2023.12

Strategy Intern (Deep Learning)

Researched point cloud characteristics of mmWave Radar and LiDAR using KITTI, Waymo and private datasets
Used GNN graph generation to synthesize diverse trajectories, output .xodr and .xosc HD map formats for simulation

🚀 Projects

Driving Simulation & Traffic Simulation Fusion PlatformSoftware Engineer · 2021.09 – 2024.06

Integrated CARLA (Unreal Engine) and SUMO traffic simulation via TCP/IP; Python scripts for vehicle control, traffic generation and data analysis
Compiled SUMO with gcc on Ubuntu; improved lane-changing model to match real-world data distributions
Used YOLOv3 for vehicle detection, trajectory extraction, and Kalman filtering for data smoothing
Real-time MySQL database for platform data with external API access

RL-Based Urban Expressway Traffic State OptimizationFirst Author · 2022.07 – 2023.04

Published SCI Q3 paper (Journal of Advanced Transportation); Q-learning for expressway ramp signal control
SUMO-Python co-simulation for coordinated ramp metering and channelization control

YOLOX-Based Road Pavement Disease DetectionTeam Lead · 2022.05 – 2022.06

Road pavement disease classification and prediction using YOLOX with data augmentation (flip, crop, rotate)
Simplified model input; improved inference speed; Excellence Award winner

📝 Publications

Deng D, Yu B, Xu D, et al. Automated Traffic State Optimization in the Weaving Area of Urban Expressways by a Reinforcement Learning-Based Cooperative Method of Channelization and Ramp Metering[J]. Journal of Advanced Transportation, 2023. (SCI Q3, IF = 2.3)

💡 Patents

Yu B, Chen Y, Deng D, Gao J, Yao S. A Traffic State Optimization Method and System in Road Interleaved Area[P]. Shanghai: CN202310359301.2, 2023-04-06.