Riassunto analitico
The homologation of Automated Driving Systems (ADS) faces significant challenges due to discrepancies between controlled track testing and real-world road conditions. Current regulatory testing frameworks struggle to comprehensively assess ADS performance across diverse driving scenarios, limiting their ability to ensure safety and reliability under variable and unpredictable conditions. This thesis addresses this gap by proposing the integration of Driver-in-the-Loop (DiL) testing into the ADS homologation process, enabling a more robust and scalable validation approach.
The proposed methodology involves developing a DiL interface for an automotive-grade static driving simulator, allowing real-time interaction between physical hardware components and virtual test environments. By incorporating dynamic traffic scenarios, sensor simulations, and decision-making algorithms, this approach enhances the representativeness of homologation tests. A specific use-case environment with realistic traffic behavior—is implemented to assess ADS perception, planning, and control strategies under mixed driving conditions.
This study details the system architecture, the development of the DiL testing framework, and the integration of regulatory test parameters into the simulation loop. Furthermore, it evaluates the feasibility of using DiL-based validation to bridge the gap between controlled track environments and open-road testing, ensuring a more comprehensive assessment of ADS safety and compliance. The findings support the adoption of DiL methodologies as a complementary tool in ADS homologation, contributing to the evolution of regulatory standards and the deployment of safer autonomous systems.
|
