SIL: Software-in-the-Loop Testing¶
SIL (Software-in-the-Loop) testing refers to a testing mode in which an autonomous-driving algorithm is compiled into an executable, connected to the simulation platform via the SimOne API, and used to validate the algorithm's control logic in a virtual environment.
The overall workflow consists of three steps:
Step 1: Add and manage the algorithm in SimOne
Package the algorithm as an .exe / .zip / Docker image, then register it in the Algorithm Management module of the SimOne front end. Both local client algorithms and cloud algorithms are supported.
→ For details, see 4.3. Algorithm Management (Algorithm Integration)
Step 2: Integrate the SimOne API and implement the algorithm control logic
The algorithm retrieves simulation data (Ego Vehicle status, sensor data, etc.) through the SimOne API and sends control commands (longitudinal/lateral control) to the simulator.
Choose the integration method based on the algorithm's development language:
- C++ algorithm integration → C++ Algorithm Integration Guide
- Python algorithm integration → Python Algorithm Integration Guide
Step 3: Create a test case and run validation
Build a test scenario in the SimOne scenario editor. When creating a run task, set the Ego Vehicle controller to API Algorithm, start the simulation, run the corresponding algorithm executable, and observe the algorithm's control behavior in the virtual environment.
For specific operations, refer to the "Test Case Editing and Running" section in the integration guides above.