4D LiDAR Self-Calibration System

Production automotive perception — online, end-of-line and after-sales calibration for solid-state LiDAR.

Role
Senior Software Developer, Product Owner, and Team Lead
Stack
C++ · Python · ROS2 · AUTOSAR · SAFe/Scrum
Status
Published
4D LiDAR Self-Calibration System — project imagery

The problem: LiDAR sensors in production vehicles need re-calibration over the product life cycle: from gradual drift, changing load states, or sensor replacement. For autonomous driving, uncorrected calibration errors degrade object detection and localization, making online-calibration a requirement for safe operation.

My Role: Building dynamic self-calibration algorithms for 4D solid-state LiDAR sensors, covering online (in-operation), dynamic end-of-line (factory), and after-sales calibration scenarios. As Product Owner I defined the full product roadmap, aligned with customers on requirements, and drove system architecture through ISO 26262 functional safety sign-off.

Technical depth:

  • AUTOSAR-compliant C++ within an in-house real-time embedded framework
  • Deployment on resource-constrained automotive embedded hardware
  • ISO 26262 / A-SPICE development; requirements via PTC Integrity / Windchill
  • Automated KPI validation via Jenkins; measurement campaign design in Python and MATLAB
ProductionAutomotive PerceptionIbeo / MicroVision 2020–2024