Jinghan Xu’s academic portfolio

About Me

I’m pursuing my Master’s degree in Transport and Geoinformation Technology at KTH. I completed my Bachelor’s degree in Civil Engineering (track: Geotechnical Engineering) at Tongji University, where I also undertook a one-year German language program and completed a minor in Artificial Intelligence. Outside of academics, I’m also a dancer with experience in choreography and freestyle battles.

Education

KTH Royal Institute of Technology (2024-08 - 2026-07)

  • Master’s Student in Transport and Geoinformation Technology.
  • KTH Scholarship Awardee.
  • Core modules: GIS and Spatial Analysis, Applied AI in Transport and GeoAI, Logistics, Statistical Learning.
  • Thesis host: WSP Sverige AB.

Tongji University (2019-09 - 2024-07)

  • Bachelor’s Degree in Civil Engineering.
  • Minor: Aritificial Intelligence; German.
  • Core modules: Mechanics, Built Environment, Applied Mathematical Modeling, Numerical Analysis.

Skills

Machine Learning & AI

  • Feature engineering (NumPy, Pandas, Scikit-learn)
  • PyTorch (training pipeline, neural network fine-tuning, customed dataset, GPU training, etc.)
  • Python for causal inference and basic statistical learning

Geospatial & Computational Engineering

  • Geospatial databases (PostgreSQL / PostGIS, PgRouting)
  • Spatial analysis and geovisualisation (ArcGIS Pro, QGIS)
  • GIS full-stack development (Mapbox GL JS)
  • Google Earth Engine, GTFS
  • Numercial analysis (OpenFOAM, PLAXIS)
  • Visualisation and postprocessing (ArcMap, AutoCAD, ParaView)

Software Development

  • C++, JavaScript, SQL
  • Qt6 (desktop GUI), React, Node.js
  • Docker (containerized development)

Languange

  • English (fluent, IELTS 7.5 / C1)
  • Mandarin (native)
  • German & French (beginner, PHS6 & TFU4 / ~A2)

Publications

  • J. Xu, Y. Ling, Z. Ma; A Generalizable Urban Green View Index Estimation Method from Satellite Imagery; Transportation Research Record (Under Review).
  • J. Xu, Y. Ling, Z. Ma; Estimating Street-Level Green View Index Using Satellite Remote Sensing and Explainable Machine Learning; TRB Annual Meeting 2026.
  • Y. Cao, J. Xu, Q. Jia, et al. Balancing the Price of Safety: A Weighted Voronoi Framework for UAM Network Design; TRB Annual Meeting 2026.
  • J. Xu, Y. Ling, Z. Ma; Convolutional Neural Network-Based Estimation of the Green View Index from Satellite Images; Swedish Transport Research Conference (STRC2025).
  • Wei H, Xu J, Jiang J, et al. Holographic display in future automotive smart cockpit: application scenarios, interaction modals, and VACP analysis; Advanced Fiber Laser Conference (AFL2022). SPIE, 2023, 12595: 200-209.

Professional Experiences

WSP Sverige AB

2026.01 - 2026.06 | Stockholm | Thesis Project Student

  • Thesis project: Planning and Optimization of Charging Infrastructure for Electric Trucks along the TEN-T Network in Sweden.

Mobility Informatics Lab, Division of Transport Planning, KTH

2025.11 - 2026.01 | Stockholm | Research Assistant

  • Participation in research on public transport contract.
  • Data evidence-based decision-making; contract design and management; operational data modeling.

Digital Futures, KTH

2025.06 - 2025.08 | Stockholm | Summer Research Intern

  • Conducted individual research project under supervision.
  • GeoAI and remote sensing for urban greenery assessment.

Innovation Center Asia, Volkswagen Group China

2024.01 - 2024.07 | Beijing | Technical Scouting Intern

  • Academic scouting for power electronics (SiC, GaN), NEV technology updates and NEV market share in China.
  • Connection and communication with related research groups in universities in China.

Innovation Center PACE, Tongji University

2022 - 2024 | Shanghai | Student Leadership Board Memeber

  • Participation and leadership in undergraduate innovation projects and competitions on mobility and transportation.
  • Connection and communication with industrial partners.

Mathart Systems Co., Ltd.

2022.01 - 2022.04 | Shanghai | Algorithm Engineer Intern

  • Coding for integer programming and heuristic algorithm modules.

Selected Projects

Rethinking public transport contracts: towards value-based KPIs for reliable bus services

Funded Research | Nov 2025 - Jan 2026 | Supervisor: Prof. Zhenliang Ma

  • Public transport contract analysis from economic perspectives: diagnosing the effectiveness of existing contracts, especially their KPIs.
  • A comprehensive review of public transport contract models: development and efficiency evaluation.
  • A data-driven proof of concept: proposing KPI negotiation using operational and GPS data.
  • This project is supported by K2 (The Swedish Knowledge Centre for Collective Mobility) and Region Stockholm (SL).

Course Project: Prioritizing Heat Risky Areas for Urban Infrastructure Improvements: A Multi-Criteria Case Study in Stockholm

Course Project (Grade: on-going) | Nov 2025 - Jan 2026 | Examiner/Supervisor: Prof. Yifang Ban

  • Multi-criteria: urban heat island effect (UHI), heat vulnerability (residence-based and mobility-based), and nature resilience.
  • Satellite imagery data collection, modeling algorithm and visualizing pipeline (python and QGIS).
  • From spatial analysis to practical conclusions: Gi* for statistically meaningful numerical boundaries, focal mean operation to include neighborhood information, DBSCAN clustering, and four-quadrant graph interpretations.
  • This project is supported by Sweco Sverige.

Coding Repository: Heat Vulnerability Analysis

Public Transport Delay Propagation Analysis: A Causal Inference Experiment

Course Project (Grade: 50/50) | Sep 2025 - Oct 2025

  • Causal inference methodology comparison for delay propagation in public transport: regression adjustment, generalized propensity score, doubly robust estimation.
  • Mobility data modeling and management: GTFS Realtime feeds, python, SQLite3.

Feel free to reach out if you’d like to read the report. Coding Repository

An AI-Based Automated Green Path Navigator & Green View Database

Course Project (Grade: A) | Sep 2025 - Oct 2025 | Examiner/Supervisor: Prof. Gyözö Gidofalvi

  • A full-stack AI-based Web GIS application: pretrained AI model, multi-modal and complex spatial data management, route planning with self-defined parameters and real-time public transport APIs.
  • Tech Stack: React, Node.js, Python (FastAPI, PyTorch), PostgreSQL/PostGIS, Docker, Mapbox GL JS, Microsoft Planetary Computer APIs, GTFS/Real-time Transit APIs.

Feel free to reach out if you’d like to read the report. Coding Repository

Cross-City Generalizable Green View Index Estimation from Satellite Imagery

Funded Research | Jun 2025 - Oct 2025 | Supervisor: Dr. Yancheng Ling; Prof. Zhenliang Ma

  • Transferrable CNN model for assessing Green View Index (GVI).
  • From data collection to model fitting: satellite imagery-derived dataset, self-defined data loaders, PyTorch training pipeline, GPU training, model deployment and performance evaluation.
  • Compared with different methods: GNN, CNN fine-tuning (ResNet), assembly (tree-based) models, regression models.
  • Explainable machine learning: SHAP, gradient attention map.
  • This work was funded by Digital Futures at KTH Royal Institute of Technology.

This work is under submission process to a journal. Coding Repository