Robert (Xiaoxuan) C. Email & Phone Number

United States

Rochelle Park, New Jersey, Us

- Developed and implemented a deep learning model (Multi-Horizon Tensor Fusion Transformer) for accurate shipment prediction. Trained the model using a large-scale dataset with billions of rows of data.- Development of LLM agents for process automation

Palo Alto, California, Us

- Use sensor fusion and computer vision algorithms for road hazard detection. - Use CNN and LSTM to label driving safety factors using time-series multi-sensor data and video embeddings.- Built a BERT-based NLP model to estimate accident costs used for company go-to-market and potential negotiation to reduce the insurance premium

Dearborn, Michigan, Us

Applied research scientist in Advanced Control Group in Ford R&A (Research and Advance Engineering Team). 2/2020 - Present- Developing a decentralized path planning algorithm for autonomous driving vehicles- Developed vehicle trajectory extraction pipeline from real-time drone video data- Filed 2 US Patents in the field of EV fleet deployment method in 2020.Full-stack research engineer focusing on Machine Learning Models and AI Applications as part of GDI&A (Global Data Insights & Analytics) 10/2018 - 2/2020- Developed and deployed a data-driven analytical solution product for the corporate decision team to visualize the model result of connected vehicle data- Collaborate with business stakeholders to solve problems which have high business value and propose analytical solutions using statistical modeling/machine learning/AI- Developed and built predictive models (incl. decision-tree-based models, neural networks, time series modeling, sequence models, reinforcement learning, etc.) to various business applications and use cases.- High dimensional data acquisition and processing, missing data completion- Published 1 US Patent, 4 journal papers in 2019.- Published 4 US Patent, 2 journal paper, and 2 conference paper in 2020.

Chicago, Illinois, Us

Data Modeling - Data storage, cleaning, processing, QAQC, natural language processing/labeling using SQL, PL/SQL, Mongo DB, ArcGIS, QGIS, R, and Python. - Zone/segment based traffic condition prediction model for the City of Chicago - Extracting real-time traffic status from large-scale XML data feed(1GB/hr) - The deploying machine learning models (CNN, ANN, RF, et al.) on human activity data (GPS, Location Based Social Media) - Automatic data integration, feature extraction, and labeling for online travel survey data. - Apply machine learning clustering method on 130 real-time traffic detectors for data quality monitoring. Data Archiving and Visualization - Archive, analysis, modeling the traffic incident data in Illinois, include freeway and local road, using tableau, python, and R. - Generate traffic and incident weekly/monthly summary report for traffic operation authorities. - Development and deployment of a data archiving system for the real-time video stream data, real-time detector data Project Managment - Assisted the project manager to create a traceability matrix captures all requirements proposed in contract, tracking the design of bus signal generator, central control software, control operation software, and bench testing.Publication and Presentation - Presented 4 conference presentations, published 6 conference papers and 3 journal papers in applying Data-Driven/Machine Learning models, human activity studies, and urban computing (2017.06 -Present) - Filed 1 US patent in advance traffic data collection and connected and automated vehicles (2017.09 -Present).

Madison, Wi, Us

- Implemented quality control criteria for traffic detector data, and created web dashboard to present the data quality of a huge database (over 4 billion records) using Java and SQL.- Statistic modeling for statewide crash data using JAVA and SAS for WI State Patrol.- Extracted statewide crash data and generated statistical analysis report for Carfax and Experian using SAS.- Optimization and prediction for transportation operation and planning using the data-driven method. - Collected, processed, and analyzed Lidar data for 3D-GIS Tech research of CV/AV.- Applied traffic flow theory to arterial traffic analysis and simulation for the City of Madison.- Published 3 conference papers and 2 journal papers.

Madison, Wi, Us

Instructed laboratory session for graduate level urban transportation planning, optimization, and traffic simulation class.

Master'S Degree, Computer Science

Self-Driving Car Engineer Program

Doctor Of Philosophy (Ph.D.), Urban Human Behavior Modeling & Intelligent Transportation System

Master Of Science (Ms), Transportation Engineering