Safety
The Safety Assurance and Cyber Security category addresses cross-functional processes and standards that need to be considered by manufacturers and developers to support the development of safe and secure ADF’s. Safety of the ADF, when undertaking the driving task within the defined ODD, is of paramount importance to the user, other road users, other traffic participants who interact with an ADF (e.g., vulnerable road users) and the reputation of manufacturers and the wider industry. To drive development towards the safe deployment of ADFs on public roads, established automotive standards e.g., ISO 26262:2018 and new standards specific to the development of more advanced driver assistance systems (ADAS) are being developed and applied.
The topics within the Safety Assurance and Cyber Security category are as follows;
- Functional Safety (ISO 26262:2018), including; identification of the hazards and risks, risk evaluation, assignment of the correct Automotive Safety Integrity Level (ASIL), definition of a safety concept for the ADF and monitoring of the safety performance of the ADF
- Safety of the intended functionality (SOTIF) (ISO 21448:2022) including; compliance with relevant standards, identification and evaluation of SOTIF related risks and the verification and validation strategy of the ADF
- Cyber Security standards, processes and best practices for automated driving functions, including; ISO/SAE 21434:2021 Road vehicles – Cybersecurity engineering, Cyber Security Management, Cyber Resilience, assessment of asset risks, hazards and mitigations, Security-informed safety (ISO/TR 4804:2020) and the validation process.
There are also relevant standards relating to Artificial Intelligence, including ISO/PAS 8800:2024 (Road Vehicles — Safety and artificial intelligence). ISO/PAS 8800:2024 defines a framework that addresses all phases of the development and deployment lifecycle of Artificial Intelligence (AI) within a road vehicle context. This is covered in more detail in the Artificial Intelligence section of this Code of Practice.
- AI – Privacy
- Requirement – Robust Design
- Requirement – Interaction with Mixed Traffic and associated Failure Modes
- Requirement – Transition of Control in Mixed Traffic
- Requirement – Interaction with Mixed Traffic and associated Risks
- Requirement – Safety Impact of Requirements
- Requirement – Risk Identification and Functional Limitation
- Requirement – Minimal Risk Manoeuvre
- Ethics – Adherence to Ethical Standards
- Functional Safety – AI and ML
- Functional Safety – Safety Concept of V2X Interactions
- Functional Safety – Minimal Risk Manoeuvre
- Functional Safety – ADF Transition to a Safe State
- Functional Safety – Derivation of Test Cases
- Functional Safety – Identification of Hazards
- Functional Safety – Safety Concept
- Functional Safety – Validation of the Safety Concept
- Functional Safety – Safety Analysis and Identification of Requirements Structure Gaps
- SOTIF – Field Monitoring Process
- SOTIF – SOTIF Release
- SOTIF – Triggering Events
- SOTIF – Limitations of the ADF
- SOTIF – Verification and Validation Strategy
- SOTIF – Driver Monitoring System
- SOTIF – Risks Relating to the Take Over Request
- SOTIF – Identification and Evaluation of Risks
- SOTIF – Functional and System Specification
- SOTIF – Safety of the Intended Functionality
- Cyber Security – Validation Concept and Process
- Cyber Security – Incident Response
- Cyber Security – Requirements Evolution through Development
- Cyber Security – Derivation of Requirements
- Cyber Security – Threat Analysis and Risk Assessment
- Cyber Security – Security by Design
- Cyber Security – Organisational Processes
- — Cross-References after this —
- Ethics – Compliance with Laws and Traffic Rules
- Ethics – Risk Assessment Compared to Human Driving
- Ethics – Compliance with Data Protection Measures
- Simulation – Analysis of ADF Impact on Simulation
- Vdata – Data Collection for Analysis of Incidents and Accidents
- Vdata – Data Protection Impact Assessment
- Vdata – Security Risk Assessment and Management for Personally Identifiable Data
- Vdata – Data Storage for EDR and DSSAD Purpose
- V2X – V2X Security Mechanisms for Vehicle Applications
- V2X – Trust Management and Misbehavior Detection
- AI – Quality of the Dataset(s) (including Training, Testing and Validation)
- AI – Safety Assessment
- AI – Identification of Safety Recommendations
- Testing – Quality of Virtual Tests
- Testing – Definition Safety Validation
- Testing – Test Concept Planning
- Testing – Test Concept Execution
- Testing – Concept Safety Validation
- Testing – Testing ODD
- Testing – Applicability of Virtual Test Tools
- System – Consideration of Architecture and External Elements
- System – Identification of Software Safety Requirements at Function Level
- System – Strategy for Ensuring Update Authenticity