Zhenyuan Ji

BMW Project - CDC(Crowd data collection) development and validation - Developed and validated CDC use cases — data collection scripts that define trigger logic, signal parsing, and sample filtering rules — for ADAS and ML data collection, covering L2 Field Monitoring, Parking Monitoring, FASSD, perception-related scenarios, and driving experience data. - Built a full validation pipeline: onboard signal triggering, data collection, byte-stream decoding, AWS backend consistency analysis, and KPI visualization. - Conducted validation across HiL test benches, prototype vehicles, and fleet testing, ensuring functional consistency at all stages. - Performed trace data analysis and backend comparison using Splunk, ensuring data consistency and anomaly detection. - Processed large-scale fleet datasets with Python, Jupyter, and AWS S3, identifying corner cases and algorithm failures. - Developed automated data quality checks for sample filtering, signal coverage monitoring, and release evaluation. - Contributed CDC use cases to the CDC Developer platform (Git-based), ensuring CI/CD consistency across teams. - Built automated reporting and visualization tools in Python, significantly improving engineering delivery efficiency. - Familiar with perception module output structures for data consistency checks and root-cause analysis.

AUDI Project - GHIL Test automation developer - Developed and maintained automated test frameworks for Audi’s GHIL platform, supporting system-level validation across multiple ECUs. - Built reusable automation modules (init/deinit, diagnostics, state handling) on Micronova EXAM, enabling efficient test case design. - Supported overnight unattended stability tests (Nachtlauf), ensuring high reliability of test automation workflows. - Configured communication and signal mapping across CAN, LIN, FlexRay, Ethernet, ensuring correct ECU integration. - Integrated Audi ODIS diagnostic flows into automated test sequences for fault injection and system-level validation. - Analyzed trace logs to locate inconsistencies and optimized model reusability and maintainability. - Maintained consistency between automated test cases and DOORS requirements for traceability and completeness.