End-to-End Closed-Loop Validation of Automated Driving (AD) Systems

Automated driving and its increasing demand have led to the surge in robustness validation of AD software in simulation. Validating multiple components of AD software in closed loop gives the opportunity to functionally validate the software in integrated setup at the simulation level.

An end-to-end closed-loop framework helps to design higher-quality products, reduce costs, and deliver innovations faster. The closed-loop validation framework helps designers to understand the cascaded impact of a module’s performance.

The major challenge in creating an end-to-end closed-loop setup is integration of multiple tools. The MathWorks product chain has helped to overcome this challenge. In this session, hear about initial development and setup for service-oriented architecture (SOA) and future work.

In this case study, an end-to-end closed-loop framework of AD systems involves these major steps:

1. Developing AD components with Simulink^® and C/C++.
2. Creating scenarios using RoadRunner and the Driving Scenario Designer app.
3. Configuring plant models and sensors.
4. Integrating AD components as Simulink modules and S-functions using Automated Driving Toolbox™.