Assisted and Self-Driving Feature Integration and Vehicle Testing
Led deployment and validation of assisted- and self-driving feature software on embedded platforms, bridging algorithm development and real-world vehicle testing for multiple OEM programs.
Project Overview
Modern driver-assistance and autonomous features rely on a blend of perception, decision, and control algorithms that begin life in high-compute PC environments.
My work focused on adapting, optimizing, and integrating these algorithms onto embedded automotive ECUs, ensuring they met real-time and memory constraints while preserving functional accuracy.
The features covered a wide spectrum—from low-speed assist functions such as Lane Keep Assist, Lane Centering, Cross-Traffic Alert, and Rear Cross-Traffic Alert, to high-speed automation like Adaptive Cruise Control and Auto Lane Change.
Once embedded deployment was complete, I led the vehicle-level testing and validation activities, working directly with algorithm, systems, and validation teams to tune performance in real driving conditions across various OEM programs and development phases.
Key Details / Outcomes
Embedded Deployment: Optimized perception and control algorithms for real-time execution on constrained hardware targets.
Feature Range: Supported both low-speed assist and high-speed highway automation features.
Vehicle Testing Leadership: Planned and executed on-road validation sessions across multiple OEM development stages.
Cross-Functional Collaboration: Served as technical liaison between algorithm developers, embedded engineers, and vehicle test teams.
Production Impact: Contributed to successful field validation and feature readiness milestones for production launches.
Challenges & Learnings
- Translating PC-based models into optimized C/C++ code suitable for automotive ECUs.
- Managing timing, memory, and sensor interface limitations in real-time environments.
- Handling dynamic behavior differences between simulated and physical vehicle tests.
- Coordinating multi-disciplinary teams across algorithm design, embedded integration, and on-road validation.
Working on assisted-driving, perception, or embedded autonomy integration?