Hardware-in-the-Loop Autonomous Driving Simulator

SelfDrivingSimulator combines hard real-time physics engines with physical automotive control electronics and dynamic human-in-the-loop motion platforms. By streaming raw GMSL2 and radar signals directly into real ECU connectors, it validates production vehicle brains in deterministic real time.

Validating Real Silicon Chips Beyond Software Pure-Virtual Abstractions

Virtual simulations running on cloud servers ignore thermal throttling, PCIe bus saturation, serializer jitter, and physical contactor bouncing. SelfDrivingSimulator places the physical autonomous computing box on a climate-controlled testbench, driving its real hardware pins with nanosecond-accurate FPGA signal injection.

Core Engineering Areas

• FPGA Raw Sensor Injection — Bypassing the camera lens by feeding pixel-level MIPI CSI-2 and GMSL2 streams directly into the vehicle computer's deserializer chips at 60 FPS.
• 6-DOF Dynamic Motion Platform — Translating simulated vehicle lateral roll, longitudinal squat, and yaw velocity into physical hydraulic and electric actuator displacements.
• Climate-Controlled Bench Integration — Subjecting the physical autonomous driving computer to -40°C to +85°C chamber temperatures while driving its power rails with voltage ripple perturbations.

Technical Articles

• Designing FPGA Direct-Injection Video Hardware for Autonomous Vehicle HIL
• Advanced Motion Cueing for High-Performance Driver-in-the-Loop Simulators

Related Cluster Architecture Specifications

• AutomotiveSDV — Full SDV Stack & Reference Architecture
• A2HCard — Agent-to-Human (A2H) Interface Cards & In-Cabin Autonomous Interaction Manifests
• A2HumanCard — Standardized In-Vehicle AI Agent Capability Discovery, Trust & Authorization Cards
• AIDrivenSynBio — AI-Driven Synthetic Biology for Automotive In-Cabin Air Purification, Bio-Engineered Pathogen Scavengers & Molecular Biosensors