SOTIF Autonomous Vehicles Safety: Evolutionary Approach

Tech News Automotive

Autonomous vehicles (AVs) aim to enhance transportation safety, efficiency, and reliability. To achieve these goals, SOTIF (Safety of the Intended Functionality) provides a comprehensive framework to address inherent performance limitations and refine complex decision-making, complementing traditional Functional Safety (FuSa) standards.

SOTIF vs Functional Safety

SOTIF (ISO 21448), also referenced in UL 4600 and ISO 34502, addresses safety risks arising from intended performance limitations and predictable misuse, rather than system failures.

Functional Safety (FuSa, ISO 26262) manages risks related to random and systematic faults.

Together, they form a comprehensive safety strategy, ensuring both technical reliability and operational robustness in complex environments.

The interaction between FUSA and SOTIF
FUSA and SOTIF scenarios. A Survey on an Emerging Safety Challenge for Autonomous Vehicles: Safety of the Intended Functionality (2024). Wang Hong, Wenbo Shao, Chen Sun

Integrating SOTIF into AV System Design

While AI-based perception, decision-making and control modules enhance situational awareness and responsiveness, they introduce challenges regarding robustness, interpretability, and edge-case performance.

By applying SOTIF (ISO 21448) principles, engineers can ensure system safety even under uncommon environmental and traffic conditions.

This structured methodology enables rigorous testing and validation across MiL, SiL and HiL, facilitating the continuous improvement and verification of autonomous functions.

Standards and Best Practices for SOTIF

Leading OEMs and Tier 1 suppliers are actively integrating SOTIF throughout the vehicle development lifecycle:

  • BMW: Implements multi-modal sensor and decision-making redundancies to mitigate inherent system limitations
  • Bosch: Employs a rigorous V-model development process specifically tailored for SOTIF, emphasizing structured verification and validation (V&V)
  • PEGASUS Project: Established a standardized 21-step methodology for the systematic assessment and validation of highly automated driving systems

Future Perspectives for SOTIF Autonomous Vehicles Safety

SOTIF represents a pivotal advancement in autonomous vehicle safety, bridging the gap between theoretical models and real-world operational robustness.

By fostering collaboration between academia, industry and regulators across the AV lifecycle, it serves as the foundational framework for safe and scalable mobility.

Ultimately, the rigorous application of SOTIF principles is the key enabler to build public trust and transform autonomous driving into a dependable everyday reality.

Written by Mario Noioso

Condividi il post

Ti potrebbe interessare anche

cover BLOG Tech news MEDIA 2

AI-Native Networks: How Artificial Intelligence Is Transforming Telecommunications

AI-Native Networks: How Artificial Intelligence Is Transforming Telecommunications Key Takeaways AI-Native Networks represent the next evolution of telecommunications infrastructure, embedding Artificial Intelligence into network operations rather than adding it as an external layer. Technologies such as AIOps, Machine Learning and Generative AI are enabling predictive maintenance, intelligent automation and faster

Leggi »
cover LUGLIO GIuseppe aprea MEDIA 2 15

Edge AI Broadcasting: When Broadcasters Stop Streaming Video

What if broadcasters stopped airing videos? Edge AI Broadcasting is redefining how media content is delivered. For decades, broadcasting has followed a simple principle: more bandwidth means better video quality. Higher bitrates, more pixels, and faster networks have driven the evolution of Media & Entertainment. But what if the next

Leggi »
Scroll to Top