[SOTIF 5-1] SOTIF's Starting Point: What Defines Safe Autonomous Driving?

Описание: What happens when an autonomous driving system gets into an accident even though nothing is broken? In this video, we explore the core concepts of ISO 21448 SOTIF (Safety of the Intended Functionality), specifically focusing on Clause 5, which defines the functional and system specifications.

While traditional Functional Safety (ISO 26262) deals with "system failures," SOTIF addresses "performance limitations" and "misuse." From blinding sunlight at a tunnel exit to Lidar in fog, and even predictable human behavior (Foreseeable Misuse), we break down how engineers identify and validate the "Unknown Unsafe" hazards.

🚗 Key Topics:
SOTIF vs. Functional Safety (ISO 26262) Differences
Physical Sensor Limitations & Triggering Conditions
Examples of Foreseeable Misuse
Validation Strategies (Simulation, Track Testing, Shadow Mode)

[Timestamps]
[00:00] Intro: The Paradox of a Zero-Failure Accident
[00:43] Physical Limitations of Sensors (Cameras and Backlighting)
[01:15] New Safety Standard: The Need for SOTIF
[01:42] Comparison of Existing Functional Safety (ISO 26262) vs. SOTIF
[02:22] Core Goal of SOTIF: Reducing Unknown Unsafe
[03:16] What is a Triggering Condition? (Example of Lidar/Radar)
[04:20] Foreseeable Misuse by Humans
[05:34] Three Verification Strategies for Proving Safety
[06:36] Residual Risk and Conclusion

#SOTIF #ISO21448 #AutonomousDriving #Clause5 #FunctionalSafety #ADAS #SafetyEngineering