{"id":198982,"date":"2026-06-04T19:34:32","date_gmt":"2026-06-04T19:34:32","guid":{"rendered":"https:\/\/innovationspace.ansys.com\/knowledge\/?post_type=topic&#038;p=198982"},"modified":"2026-06-04T19:34:52","modified_gmt":"2026-06-04T19:34:52","slug":"safety-of-ai-ml-systems-in-road-vehicles-compliance-with-iso-pas-8800","status":"publish","type":"topic","link":"https:\/\/innovationspace.ansys.com\/knowledge\/forums\/topic\/safety-of-ai-ml-systems-in-road-vehicles-compliance-with-iso-pas-8800\/","title":{"rendered":"Safety of AI\/ML systems in road vehicles &#8211; Compliance with ISO\/PAS 8800"},"content":{"rendered":"<p style=\"text-align: center\">\n    <img decoding=\"async\" src=\"https:\/\/innovationspace.ansys.com\/knowledge\/wp-content\/uploads\/sites\/4\/2026\/06\/scade-063-banner.jpeg\" style=\"max-height: 700px !important\" \/><br \/>\n    <em><\/em>\n<\/p>\n<p>This article is part of a blog series. Here are links to each part:<\/p>\n<ul>\n<li><strong>Part 1<\/strong>: <a href=\"https:\/\/innovationspace.ansys.com\/knowledge\/forums\/topic\/towards-trustworthy-ai-ml-based-systems\/\">Towards trustworthy AI\/ML-based systems<\/a><\/li>\n<li><strong>Part 2<\/strong> (this blog): Safety of AI\/ML Systems in Road Vehicles &#8211; Compliance with ISO\/PAS 8800<\/li>\n<\/ul>\n<p>This blog series explores how artificial intelligence (AI) and machine learning (ML) are reshaping autonomy, systems engineering, software development, and certification in safety-critical domains. Its goal is to help engineers and decision-makers navigate a fast-evolving landscape of recommended practices, emerging standards, engineering methods, and real-world applications.<\/p>\n<p>In Part 1, we looked at the broader challenge of building trustworthy AI\/ML-based systems. In this second article, the focus shifts to a concrete automotive safety framework: ISO\/PAS 8800, published in December 2024 to provide guidance for the safe integration of AI\/ML technologies into road-vehicle systems. We summarize the purpose and structure of the standard, explain how it complements established references such as ISO 26262 and ISO 21448, and highlight how model-based engineering, simulation, and safety analysis can support compliance activities across the AI\/ML lifecycle.<\/p>\n<h3  id=\"WHAT-IS-ISO-PAS-8800\">What is ISO\/PAS 8800?<\/h3>\n<p><a href=\"https:\/\/www.iso.org\/standard\/83303.html\">ISO\/PAS 8800<\/a> was published in December 2024 by the International Standardization Organization (ISO) to provide guidelines for safety of AI\/MLs technology when integrated into road-vehicle systems. The specification includes prerequisites, objectives and requirements for design and development of Electric and Electronic (E\/E) Systems integrating AI\/ML components in the automotive domain covering all phases of the so-called safety AI\/ML lifecycle. This specification is aligned and need to be used in conjunction with existing standards, e.g., <a href=\"https:\/\/www.iso.org\/publication\/PUB200262.html\">ISO\/IEC 26262<\/a> (Functional Safety), and <a href=\"https:\/\/www.iso.org\/standard\/77490.html\">ISO\/IEC 21448<\/a> (Safety of the Intended Function).<\/p>\n<h3  id=\"WHY-AND-WHEN-IS-ISO-PAS-8800-NEEDED\">Why and when is ISO\/PAS 8800 needed?<\/h3>\n<p>The introduction of AI\/ML technology faces challenges and shall lead to changes in the way automotive systems are designed and developed. Reference standards for safety in the automotive domain, <a href=\"https:\/\/www.iso.org\/publication\/PUB200262.html\">ISO\/IEC 26262<\/a>, and <a href=\"https:\/\/www.iso.org\/standard\/77490.html\">ISO\/IEC 21448<\/a>, are limited when addressing specifics of AI\/ML algorithms, namely:<\/p>\n<ul>\n<li>Categories of AI\/ML errors,<\/li>\n<li>Sufficiency of AI\/ML component performance,<\/li>\n<li>Impacts of AI\/ML components on safety and hazardous situations,  <\/li>\n<li>Safety Lifecycle for AI\/ML design and development. <\/li>\n<\/ul>\n<p>ISO\/PAS 8800 shall be applied whenever an AI\/ML-based System needs to satisfy safety requirements ensuring operation as expected, within its foreseen Operating Environment (or Operational Design Domain), thus supporting its intended functionality by ensuring AI\/ML components deliver outcomes according to allocated requirements for: <\/p>\n<ul>\n<li>Acceptable AI\/ML performance,<\/li>\n<li>Output quality considering AI\/ML errors,<\/li>\n<li>Limiting impacts on hazardous events at vehicle level.<\/li>\n<\/ul>\n<h3  id=\"AI-ML-SPECIFICS-FOR-ROAD-VEHICLES-SAFETY\">AI\/ML Specifics for Road-vehicles Safety<\/h3>\n<p>Integration of AI\/ML technology in the automotive domain exhibits the following salient traits:<\/p>\n<p><strong>Increased autonomy levels.<\/strong> A vast majority of projects integrating AI\/ML technology in automotive target medium to high levels of autonomy (L2+ to L5). Autonomy becomes a factor during AI\/ML integration.<\/p>\n<p><strong>Acceptability of AI\/ML errors.<\/strong> The integration of AI\/ML algorithms requires a comprehensive demonstration of the AI\/ML errors contributing to overall system performance and their acceptability determined by their impact on safety.<\/p>\n<p><strong>Functional gain and affordability.<\/strong> Projects that integrate AI\/ML technology require identification of the functional gain and its affordability. This requires having viable timelines and costs resulting from addressing specifics of AI\/ML systems design and development, e.g., adapting existing lifecycles and existing know-how.<\/p>\n<h3  id=\"KEY-COMPONENTS-OF-ISO-PAS-8800\">Key components of ISO\/PAS 8800<\/h3>\n<p><em>Clause 6 &#8211; AI within the context of road vehicles system safety engineering and basic concepts.<\/em> Along with scope and terminology, this clause positions ISO\/PAS 8800 in regard to existing safety standards, namely <a href=\"https:\/\/www.iso.org\/publication\/PUB200262.html\">ISO 26262<\/a> and <a href=\"https:\/\/www.iso.org\/standard\/77490.html\">ISO 21448<\/a>. It also provides a first classification for AI\/ML error categories.<\/p>\n<p><em>Clause 7 &#8211; AI safety management.<\/em> It defines the safety AI\/ML lifecycle for design and development, starting at encompassing system level. The clause incorporates an iterative V-cycle covering design, development, validation and verification of the AI\/ML system and component(s), including data-related considerations. The AI\/ML lifecycle covers integration of AI\/ML algorithms into the encompassing system and operation in real context, as shown in the figure below (borrowed from <a href=\"https:\/\/www.iso.org\/standard\/83303.html\">ISO\/PAS 8800<\/a>).<\/p>\n<p style=\"text-align: center\">\n    <img decoding=\"async\" src=\"https:\/\/innovationspace.ansys.com\/knowledge\/wp-content\/uploads\/sites\/4\/2026\/06\/scade-063-ai-system-design-cycle.png\" style=\"max-height: 350px !important\" \/><br \/>\n    <em><\/em>\n<\/p>\n<p><em>Clause 8 &#8211; Assurance arguments for AI systems.<\/em> A distinctive trait of ISO\/PAS 8800 is that the AI\/ML lifecycle begins with assurance argumentation. This phase shall identify all AI\/ML specifics, considering the context of usage, categories of evidence, and roots of truth to anchor evidence. The assurance argumentation supports the structure for demonstration of safety (i.e., the safety case). It also provides a foundation for requirements to be elicited and to ensure quality attributes such as correctness, sufficiency, and consistency.<\/p>\n<p><em>Clause 9 &#8211; Derivation of AI safety requirements.<\/em> Derivation of AI\/ML requirements can follow a combined top-down, bottom-up approach. Regarding top-down, existing safety techniques, for instance HARA, FMEA, and FTA, can be tailored from encompassing system level towards AI\/ML component level to derive requirements incorporating system level considerations (e.g., relying upon FuSa and SOTIF analyses). As for bottom-up, AI\/ML related properties, namely generalization, robustness, explainability, etc. are considered to determine their applicability during requirements derivation. Derivation of thresholds in AI\/ML requirements specification is of utmost importance.<\/p>\n<p><em>Clause 10 &#8211; Selection of AI technologies, architectural and development measures.<\/em> It provides recommendations for selection of AI\/ML technologies, considering the features of each AI\/ML category and which properties they contribute to support. It also recommends deployment of measures to increase confidence in the usage of AI\/ML technology according to known errors or limitations.<\/p>\n<p><em>Clause 11 &#8211; Data-related considerations.<\/em> This clause contains a specific lifecycle dedicated to dataset development. The dataset lifecycle covers different phases: from safety analysis, requirements, design, implementation, verification, validation, up to maintenance. A comprehensive list of data-related properties is provided to be considered during requirement derivation, e.g., representativeness, fidelity, completeness, traceability, etc.<\/p>\n<p><em>Clause 12 &#8211; Verification and validation of the AI system.<\/em> It includes reference processes to conduct verification and validation of the AI\/ML system and components. A variety of testing methods is suggested to conduct validation including, but not limited to, synthetic test case generation, expert knowledge, robustness testing, etc. Recommendations are also provided regarding AI\/ML integration and verification, including applicable techniques, like probabilistic testing, scenario-based testing and usage of synthetic data environments. <\/p>\n<p><em>Clause 13 &#8211; Safety analysis of AI system.<\/em> This clause focuses on techniques that can be applied or adapted to analyze safety of the AI\/ML system. The techniques in scope comprise, but are not limited to, FTA, FMEA, STPA, ETA, HAZOP, Bayesian networks, etc.<\/p>\n<p><em>Clause 14  &#8211;  Measures during operation.<\/em> Measures during operation need to be identified to achieve continual, periodic re-evaluation of the assurance arguments (safety case). This activity shall assess remaining uncertainties in the implementation, unveil unexpected operation behavior, and facilitate early identification of evolutions in the Operational Design Domain (e.g., data drifts).  This phase of the AI\/ML lifecycle may include periodically collecting data and information from the AI\/ML system operation (e.g., via monitoring), and analyzing collected data (e.g., incident analysis). Technical safety measures are elicited accordingly to comply with safety argumentation (e.g., retraining the AI\/ML algorithm).<\/p>\n<p><em>Clause 15  &#8211;  Confidence in use of AI development frameworks and software tools used for AI model development.<\/em> This clause introduces processes to increase confidence in tools used in the AI\/ML lifecycle including both traditional software tools and AI-based tools, with a focus on data-related processes. This clause shall still be consolidated by upcoming standards, e.g., <a href=\"https:\/\/www.iso.org\/standard\/89535.html\">ISO 22440<\/a>.  <\/p>\n<h3  id=\"SUPPORT-FOR-COMPLIANCE-TO-ISO-PAS-8800\">Support for Compliance to ISO\/PAS 8800<\/h3>\n<p>Synopsys develops a Digital Engineering \/ Model Based Systems Engineering (DEM) methodology to support phases and activities in the reference Safety AI\/ML lifecycle in ISO\/PAS 8800. The DEM methodology relies upon a collection of tools that can be composed, forming workflows that facilitate engineering tasks. An overview of the tools is presented below.<\/p>\n<p style=\"text-align: center\">\n    <img decoding=\"async\" src=\"https:\/\/innovationspace.ansys.com\/knowledge\/wp-content\/uploads\/sites\/4\/2026\/05\/scade-063-ansys-autonomy-solution-1.png\" style=\"max-height: 500px !important\" \/><br \/>\n    <em><\/em>\n<\/p>\n<p>The related work products constitute evidence for compliance. Tool support for different activities per clause is detailed in the following items.<\/p>\n<p><strong>Clause 7, &#8220;<em>AI safety management<\/em>&#8220;<\/strong> <\/p>\n<table style=\"max-width: 1000px\">\n<tr>\n<td style=\"padding: 0px 10px;min-width: 150px;text-align: center\">\n<p><strong>Tool<\/strong><\/p>\n<\/td>\n<td style=\"padding: 0px 10px;min-width: 150px;text-align: center\">\n<p><strong>Supported Activities for ISO\/PAS 8800 Compliance<\/strong><\/p>\n<\/td>\n<\/tr>\n<tr>\n<td style=\"padding: 0px 10px;min-width: 150px\">\n<p><a href=\"https:\/\/www.ansys.com\/fr-fr\/products\/connect\/ansys-system-architecture-modeler\">SAM<\/a><\/p>\n<\/td>\n<td style=\"padding: 0px 10px;min-width: 150px\">\n<ul>\n<li>Capturing prerequisites of the AI safety management process, namely safety requirements, AI system functionality and validation targets (acc. ISO 26262:2018 Part 3 and Part 4, and ISO 21448:2022 &#8211; Part 6 and Part 9)<\/li>\n<li>Specification of the lifecycle process, its activities, work products and related outcomes.<\/li>\n<\/ul>\n<\/td>\n<\/tr>\n<tr>\n<td style=\"padding: 0px 10px;min-width: 150px\">\n<p><a href=\"https:\/\/www.ansys.com\/products\/safety-analysis\/ansys-digital-safety-manager\">DSM<\/a><\/p>\n<\/td>\n<td style=\"padding: 0px 10px;min-width: 150px\">\n<ul>\n<li>Substantiation of safety case via incorporation of links to outcomes and work products from activities in the AI safety lifecycle process.<\/li>\n<\/ul>\n<\/td>\n<\/tr>\n<\/table>\n<p><strong>Clause 8<\/strong>, <strong>&#8220;<em>Assurance Arguments for AI Systems<\/em>&#8220;<\/strong> <\/p>\n<table style=\"max-width: 1000px\">\n<tr>\n<td style=\"padding: 0px 10px;min-width: 150px;text-align: center\">\n<p><strong>Tool<\/strong><\/p>\n<\/td>\n<td style=\"padding: 0px 10px;min-width: 150px;text-align: center\">\n<p><strong>Supported Activities for ISO\/PAS 8800 Compliance<\/strong><\/p>\n<\/td>\n<\/tr>\n<tr>\n<td style=\"padding: 0px 10px;min-width: 150px\">\n<p><a href=\"https:\/\/www.ansys.com\/products\/safety-analysis\/ansys-digital-safety-manager\">DSM<\/a><\/p>\n<\/td>\n<td style=\"padding: 0px 10px;min-width: 150px\">\n<ul>\n<li>Capturing and analyzing the assurance arguments in Goal Standard Notation for the encompassing and AI systems according to safety requirements<em>.<\/em><\/li>\n<li>Incorporation of traceability links to the confirmation measures to be applied, according to the assurance arguments, facilitating measures scheduling and management (e.g., confirmation after incorporation).<\/li>\n<\/ul>\n<\/td>\n<\/tr>\n<tr>\n<td style=\"padding: 0px 10px;min-width: 150px\">\n<p><a href=\"https:\/\/www.ansys.com\/fr-fr\/products\/connect\/ansys-system-architecture-modeler\">SAM<\/a><\/p>\n<\/td>\n<td style=\"padding: 0px 10px;min-width: 150px\">\n<ul>\n<li>Including requirements and other design and safety-related artifacts (e.g., AI system architecture, incorporated AI-safety measures) to substantiate the safety case.<\/li>\n<li>Incorporation of traceability links to safety-related artifacts to substantiate the safety case.<\/li>\n<\/ul>\n<\/td>\n<\/tr>\n<\/table>\n<p><strong>Clause 9, &#8220;<em>Derivation of Safety Requirements<\/em>&#8220;<\/strong> <\/p>\n<table style=\"max-width: 1000px\">\n<tr>\n<td style=\"padding: 0px 10px;min-width: 150px;text-align: center\">\n<p><strong>Tool<\/strong><\/p>\n<\/td>\n<td style=\"padding: 0px 10px;min-width: 150px;text-align: center\">\n<p><strong>Supported Activities for ISO\/PAS 8800 Compliance<\/strong><\/p>\n<\/td>\n<\/tr>\n<tr>\n<td style=\"padding: 0px 10px;min-width: 150px\">\n<p><a href=\"https:\/\/www.ansys.com\/fr-fr\/products\/connect\/ansys-system-architecture-modeler\">SAM<\/a><\/p>\n<\/td>\n<td style=\"padding: 0px 10px;min-width: 150px\">\n<ul>\n<li>Requirements modelling, their refinements, derivations, and allocations, from the encompassing system towards the AI system, including operational conditions and the input space of the AI system. <\/li>\n<li>Requirements traceability to identify non-fulfilled AI safety requirements, including those related to the input space of the AI system, and to trace them back to their sources.<\/li>\n<\/ul>\n<\/td>\n<\/tr>\n<tr>\n<td style=\"padding: 0px 10px;min-width: 150px\">\n<p><a href=\"https:\/\/www.ansys.com\/products\/safety-analysis\/ansys-digital-safety-manager\">DSM<\/a><\/p>\n<\/td>\n<td style=\"padding: 0px 10px;min-width: 150px\">\n<ul>\n<li>Specification of arguments in Goal Standard Notation to justify the refined\/derived requirements are adequate to fulfill AI safety requirements,<\/li>\n<li>Analysis to prevent\/control AI system functional insufficiencies.<\/li>\n<\/ul>\n<\/td>\n<\/tr>\n<\/table>\n<p><strong>Clause 10, &#8220;<em>Selection of AI technologies, architectural and development measures<\/em>&#8220;<\/strong> <\/p>\n<table style=\"max-width: 1000px\">\n<tr>\n<td style=\"padding: 0px 10px;min-width: 150px;text-align: center\">\n<p><strong>Tool<\/strong><\/p>\n<\/td>\n<td style=\"padding: 0px 10px;min-width: 150px;text-align: center\">\n<p><strong>Supported Activities for ISO\/PAS 8800 Compliance<\/strong><\/p>\n<\/td>\n<\/tr>\n<tr>\n<td style=\"padding: 0px 10px;min-width: 150px\">\n<p><a href=\"https:\/\/www.ansys.com\/products\/safety-analysis\/ansys-digital-safety-manager\">DSM<\/a><\/p>\n<\/td>\n<td style=\"padding: 0px 10px;min-width: 150px\">\n<ul>\n<li>Capturing the argumentation for AI safety requirement fulfilment in Goal Standard Notation.<\/li>\n<\/ul>\n<\/td>\n<\/tr>\n<tr>\n<td style=\"padding: 0px 10px;min-width: 150px\">\n<p><a href=\"https:\/\/www.ansys.com\/fr-fr\/products\/connect\/ansys-system-architecture-modeler\">SAM<\/a><\/p>\n<\/td>\n<td style=\"padding: 0px 10px;min-width: 150px\">\n<ul>\n<li>Modeling the allocation of AI safety requirements towards AI components,<\/li>\n<li>Traceability for validation of architectural measures, incorporated to reduce risks related to AI errors or AI insufficiencies, to fulfill requirements.<\/li>\n<\/ul>\n<\/td>\n<\/tr>\n<tr>\n<td style=\"padding: 0px 10px;min-width: 150px\">\n<p><a href=\"https:\/\/www.ansys.com\/products\/safety-analysis\/ansys-medini-analyze\">medini analyze<\/a><\/p>\n<\/td>\n<td style=\"padding: 0px 10px;min-width: 150px\">\n<ul>\n<li>Risks\/safety analyses relying upon HARA, FMEA, FTA, and STPA starting at encompassing system level.<\/li>\n<\/ul>\n<\/td>\n<\/tr>\n<tr>\n<td style=\"padding: 0px 10px;min-width: 150px\">\n<p><a href=\"https:\/\/www.ansys.com\/products\/embedded-software\/ansys-scade-suite\">SCADE Suite<\/a><\/p>\n<\/td>\n<td style=\"padding: 0px 10px;min-width: 150px\">\n<ul>\n<li>Together with the <a href=\"https:\/\/developer.ansys.com\/app\/ansys-inc\/neural-networks-package\/1\">ONXX-SCADE importer<\/a>, a back-to-back performance analysis helps to identify differences between ML Models in development and target environments, as well as impact on safety requirements.<\/li>\n<\/ul>\n<\/td>\n<\/tr>\n<tr>\n<td style=\"padding: 0px 10px;min-width: 150px\">\n<p><a href=\"https:\/\/www.ansys.com\/products\/av-simulation\/ansys-avxcelerate-sensors\">AVxcelerate Sensors<\/a><\/p>\n<\/td>\n<td style=\"padding: 0px 10px;min-width: 150px\">\n<ul>\n<li>Generating synthetic data to fulfill requirements related to the Operational Design Domain and the input space of the AI system, e.g., generation of datasets used for ML training, validation, and testing.<\/li>\n<\/ul>\n<\/td>\n<\/tr>\n<tr>\n<td style=\"padding: 0px 10px;min-width: 150px\">\n<p><a href=\"https:\/\/www.ansys.com\/products\/connect\/ansys-optislang\">optiSLang<\/a><\/p>\n<\/td>\n<td style=\"padding: 0px 10px;min-width: 150px\">\n<ul>\n<li>Generating concrete data samples for Operational Design Domain and input space exploration, in compliance with the probability distributions specified in requirements.<\/li>\n<li>Ensuring representativeness of AI\/ML input space by ensuring data are in compliance with the true\/theoretical probability distributions.<\/li>\n<\/ul>\n<\/td>\n<\/tr>\n<\/table>\n<p><strong>Clause 11<\/strong>, <strong>&#8220;Data-related considerations&#8221;<\/strong> <\/p>\n<table style=\"max-width: 1000px\">\n<tr>\n<td style=\"padding: 0px 10px;min-width: 150px;text-align: center\">\n<p><strong>Tool<\/strong><\/p>\n<\/td>\n<td style=\"padding: 0px 10px;min-width: 150px;text-align: center\">\n<p><strong>Supported Activities for ISO\/PAS 8800 Compliance<\/strong><\/p>\n<\/td>\n<\/tr>\n<tr>\n<td style=\"padding: 0px 10px;min-width: 150px\">\n<p><a href=\"https:\/\/www.ansys.com\/fr-fr\/products\/connect\/ansys-system-architecture-modeler\">SAM<\/a><\/p>\n<\/td>\n<td style=\"padding: 0px 10px;min-width: 150px\">\n<ul>\n<li>Specifying the AI dataset lifecycle process supporting iterative development, and requirements updates, e.g., after processing AI\/ML insufficiencies. <\/li>\n<li>Tracing and documenting activities related to datasets like gathering, creation, safety analysis, verification, validation, management, and maintenance of datasets.<\/li>\n<\/ul>\n<\/td>\n<\/tr>\n<tr>\n<td style=\"padding: 0px 10px;min-width: 150px\">\n<p><a href=\"https:\/\/www.ansys.com\/products\/safety-analysis\/ansys-medini-analyze\">medini analyze<\/a><\/p>\n<\/td>\n<td style=\"padding: 0px 10px;min-width: 150px\">\n<ul>\n<li>The Process FMEA feature supports <strong>data-related safety analyses<\/strong>, e.g., <strong>Data-driven FMEA<\/strong>.<\/li>\n<\/ul>\n<\/td>\n<\/tr>\n<tr>\n<td style=\"padding: 0px 10px;min-width: 150px\">\n<p><a href=\"https:\/\/www.ansys.com\/products\/av-simulation\/ansys-avxcelerate-sensors\">AVx Sensors<\/a><\/p>\n<\/td>\n<td style=\"padding: 0px 10px;min-width: 150px\">\n<ul>\n<li>Generation of synthetic data based upon different sensor technology: Camera, LiDAR, RADAR, Thermal.<\/li>\n<li>Sourcing data for scenario-based testing.<\/li>\n<li>Sourcing data for AI\/ML datasets used for training, validation and testing.<\/li>\n<\/ul>\n<\/td>\n<\/tr>\n<\/table>\n<p><strong>Clause 12, &#8220;Verification and Validation of the AI system&#8221;<\/strong> <\/p>\n<table style=\"max-width: 1000px\">\n<tr>\n<td style=\"padding: 0px 10px;min-width: 150px;text-align: center\">\n<p><strong>Tool<\/strong><\/p>\n<\/td>\n<td style=\"padding: 0px 10px;min-width: 150px;text-align: center\">\n<p><strong>Supported Activities for ISO\/PAS 8800 Compliance<\/strong><\/p>\n<\/td>\n<\/tr>\n<tr>\n<td style=\"padding: 0px 10px;min-width: 150px\">\n<p><a href=\"https:\/\/www.ansys.com\/products\/av-simulation\/ansys-avxcelerate-autonomy\">AVxcelerate Autonomy<\/a><\/p>\n<\/td>\n<td style=\"padding: 0px 10px;min-width: 150px\">\n<ul>\n<li>Validation and Verification of the encompassing system and the AI system requirements.<\/li>\n<li>Generation of test cases and variations for scenario-based testing (different testing methods supported).<\/li>\n<li>Incorporation of algorithms for validation of properties, e.g., for stability, generalization, and robustness.<\/li>\n<li>Analysis of coverage of the AI system input space according to predefined KPIs.<\/li>\n<li>Verification of the AI-safety requirements allocated to the AI component.<\/li>\n<\/ul>\n<\/td>\n<\/tr>\n<\/table>\n<p><strong>Clause 13, &#8220;Safety Analysis of AI systems&#8221;<\/strong><\/p>\n<table style=\"max-width: 1000px\">\n<tr>\n<td style=\"padding: 0px 10px;min-width: 150px;text-align: center\">\n<p><strong>Tool<\/strong><\/p>\n<\/td>\n<td style=\"padding: 0px 10px;min-width: 150px;text-align: center\">\n<p><strong>Supported Activities for ISO\/PAS 8800 Compliance<\/strong><\/p>\n<\/td>\n<\/tr>\n<tr>\n<td style=\"padding: 0px 10px;min-width: 150px\">\n<p><a href=\"https:\/\/www.ansys.com\/products\/safety-analysis\/ansys-medini-analyze\">medini analyze<\/a><\/p>\n<\/td>\n<td style=\"padding: 0px 10px;min-width: 150px\">\n<ul>\n<li>Tailoring typical techniques like FMEA, HARA, FTA and STPA to analyze AI errors at encompassing and AI systems levels, assuming the AI\/ML component as a black box, <\/li>\n<li>Definition of the mitigation measures to cope with AI\/ML system errors leading to unbearable risks.<\/li>\n<\/ul>\n<\/td>\n<\/tr>\n<tr>\n<td style=\"padding: 0px 10px;min-width: 150px\">\n<p><a href=\"https:\/\/www.ansys.com\/fr-fr\/products\/connect\/ansys-system-architecture-modeler\">SAM<\/a><strong> &amp; <\/strong><a href=\"https:\/\/www.ansys.com\/products\/safety-analysis\/ansys-medini-analyze\">medini analyze<\/a><\/p>\n<\/td>\n<td style=\"padding: 0px 10px;min-width: 150px\">\n<ul>\n<li>Traceability between <strong>SAM<\/strong> and <strong>medini analyze<\/strong> ensures that the requirements derived after safety analyses are traceable from the architecture model towards <strong>SAM<\/strong> whenever required, e.g., to incorporate an architectural measure or update requirements.<\/li>\n<\/ul>\n<\/td>\n<\/tr>\n<\/table>\n<p><strong>Clause 14, &#8220;Measures during operation&#8221;<\/strong><\/p>\n<table style=\"max-width: 1000px\">\n<tr>\n<td style=\"padding: 0px 10px;min-width: 150px;text-align: center\">\n<p><strong>Tool<\/strong><\/p>\n<\/td>\n<td style=\"padding: 0px 10px;min-width: 150px;text-align: center\">\n<p><strong>Supported Activities for ISO\/PAS 8800 Compliance<\/strong><\/p>\n<\/td>\n<\/tr>\n<tr>\n<td style=\"padding: 0px 10px;min-width: 150px\">\n<p><a href=\"https:\/\/www.ansys.com\/products\/safety-analysis\/ansys-medini-analyze\">medini analyze<\/a><\/p>\n<\/td>\n<td style=\"padding: 0px 10px;min-width: 150px\">\n<ul>\n<li>Analysis of safety risks associated to the AI\/ML system operation.<\/li>\n<\/ul>\n<\/td>\n<\/tr>\n<tr>\n<td style=\"padding: 0px 10px;min-width: 150px\">\n<p><a href=\"https:\/\/www.ansys.com\/fr-fr\/products\/connect\/ansys-system-architecture-modeler\">SAM<\/a><\/p>\n<\/td>\n<td style=\"padding: 0px 10px;min-width: 150px\">\n<ul>\n<li>Specification of derived requirements and associated measures to cope with safety risks. <\/li>\n<li>Specification of measures at the encompassing System or AI\/ML system levels.<\/li>\n<\/ul>\n<\/td>\n<\/tr>\n<tr>\n<td style=\"padding: 0px 10px;min-width: 150px\">\n<p><a href=\"https:\/\/www.ansys.com\/products\/safety-analysis\/ansys-digital-safety-manager\">DSM<\/a><\/p>\n<\/td>\n<td style=\"padding: 0px 10px;min-width: 150px\">\n<ul>\n<li>Incorporation of traceability links for measures. <\/li>\n<li>Traceability for new safety requirements and measures for AI\/ML system operation.<\/li>\n<li>Specification\/traceability of evidence to substantiate the safety case.<\/li>\n<\/ul>\n<\/td>\n<\/tr>\n<\/table>\n<p><strong>Clause 15, &#8220;Confidence in use of AI development frameworks and software tools used for AI model development&#8221;<\/strong><\/p>\n<table style=\"max-width: 1000px\">\n<tr>\n<td style=\"padding: 0px 10px;min-width: 150px;text-align: center\">\n<p><strong>Tool<\/strong><\/p>\n<\/td>\n<td style=\"padding: 0px 10px;min-width: 150px;text-align: center\">\n<p><strong>Supported Activities for ISO\/PAS 8800 Compliance<\/strong><\/p>\n<\/td>\n<\/tr>\n<tr>\n<td style=\"padding: 0px 10px;min-width: 150px\">\n<p><a href=\"https:\/\/www.ansys.com\/products\/safety-analysis\/ansys-medini-analyze\">medini analyze<\/a><\/p>\n<\/td>\n<td style=\"padding: 0px 10px;min-width: 150px\">\n<ul>\n<li>Qualified TCL 3 up to ASIL D, acc. ISO 26262.<\/li>\n<\/ul>\n<\/td>\n<\/tr>\n<tr>\n<td style=\"padding: 0px 10px;min-width: 150px\">\n<p><a href=\"https:\/\/www.ansys.com\/products\/embedded-software\/ansys-scade-suite\">SCADE Suite<\/a><\/p>\n<\/td>\n<td style=\"padding: 0px 10px;min-width: 150px\">\n<ul>\n<li>Qualified TCL 3 up to ASIL D, acc. ISO 26262.<\/li>\n<\/ul>\n<\/td>\n<\/tr>\n<tr>\n<td style=\"padding: 0px 10px;min-width: 150px\">\n<p><a href=\"https:\/\/www.ansys.com\/products\/av-simulation\/ansys-avxcelerate-autonomy\">AVxcelerate Autonomy<\/a><\/p>\n<\/td>\n<td style=\"padding: 0px 10px;min-width: 150px\">\n<ul>\n<li>Qualified TCL 3 up to ASIL D, acc. ISO 26262.<\/li>\n<\/ul>\n<\/td>\n<\/tr>\n<\/table>\n<h4  id=\"ANSYS-SUITE-CORE-TOOLS\">Ansys Suite Core Tools<\/h4>\n<p>The distinctive features of the Ansys suite are summarized inline:<\/p>\n<ul>\n<li><a href=\"https:\/\/www.ansys.com\/products\/safety-analysis\/ansys-medini-analyze\">medini analyze<\/a><em>:<\/em> Systems, HW and SW modeler implementing analysis methods for safety (HAZOP, HARA, FHA, FTA, FMEA) and cyber-security (TOE, Attack Tree, TARA) according to standards.<\/li>\n<li><a href=\"https:\/\/www.ansys.com\/fr-fr\/products\/connect\/ansys-system-architecture-modeler\">SAM<\/a><em>:<\/em> <a href=\"https:\/\/www.omg.org\/spec\/SysML\/2.0\/Beta1\/Language\/PDF\">SysML v2<\/a> modeler for Systems Architecture, Use Case, Activity and Requirements engineering.<\/li>\n<li><a href=\"https:\/\/www.ansys.com\/products\/embedded-software\/ansys-scade-suite\">SCADE<\/a><em>:<\/em> Environment for reliable and safe embedded SW modeling, verification, and code generation, compliant with aeronautics, automotive, railway, nuclear and general industries safety standards.<\/li>\n<li><a href=\"https:\/\/www.ansys.com\/products\/safety-analysis\/ansys-digital-safety-manager\">DSM<\/a><em>:<\/em> The Digital Safety Manager drives optimization of the safety-process acting as a central hub to gather data, managing resources, planning, etc. for systems, HW and SW development projects.<\/li>\n<li><a href=\"https:\/\/www.ansys.com\/products\/av-simulation\/ansys-avxcelerate-sensors\">AVx Sensors<\/a><em>:<\/em> Simulation engine including a catalogue for sensors-perception simulation and capabilities to evaluate Autonomous systems.<\/li>\n<li><a href=\"https:\/\/www.ansys.com\/products\/connect\/ansys-optislang\">optiSlang<\/a><em>:<\/em> Environment for parametric designs studies, process integration &amp; automation. Provides AI\/ML and non-AI\/ML based algorithms for e.g. Sensitivity study, Optimization or Robust Design and connects those to tool chains of engineering tools. <\/li>\n<li><a href=\"https:\/\/www.ansys.com\/products\/av-simulation\/ansys-avxcelerate-autonomy\">AVxcelerate Autonomy<\/a><em>:<\/em> Cloud-based simulator, including compositional workflow and interfaces with optiSlang and AVx Sensors, able to evaluate autonomous Systems\/SW performance and safety-related indicators based upon Open-Scenario format.<\/li>\n<\/ul>\n<h3  id=\"CONCLUSION\">Conclusion<\/h3>\n<p>As AI\/ML capabilities move deeper into road-vehicle systems, safety assurance can no longer rely only on extending traditional automotive processes. It also requires methods that explicitly address AI\/ML-specific challenges, including data dependence, performance uncertainty, validation coverage, and the need for ongoing operational monitoring.<\/p>\n<p>ISO\/PAS 8800 provides a structured way to address these challenges while remaining aligned with established standards such as ISO 26262 and ISO 21448. Its value lies in extending the safety framework to cover assurance arguments, AI\/ML-specific requirements, data lifecycle activities, verification and validation, safety analysis, and measures during operation.<\/p>\n<p>Combined with model-based engineering, simulation, safety analysis, and end-to-end traceability, ISO\/PAS 8800 offers a practical foundation for building justified confidence in AI\/ML-enabled automotive systems.<\/p>\n<p>To learn more about Synopsys embedded software and digital engineering solutions for safety-critical development, get in touch on our <a href=\"https:\/\/www.ansys.com\/contact-us\">contact page<\/a>.<\/p>\n<h3  id=\"ABOUT-THE-AUTHOR\">About the author<\/h3>\n<table style=\"max-width: 1000px;border: none !important\">\n<tr>\n<td style=\"padding: 0px 10px;min-width: 150px;border: none !important\">\n<p style=\"text-align: center\">\n    <img decoding=\"async\" src=\"https:\/\/innovationspace.ansys.com\/knowledge\/wp-content\/uploads\/sites\/4\/2026\/06\/scade-063-author.png\" style=\"max-height: 150px !important\" \/><br \/>\n                <em><\/em>\n<\/p>\n<\/td>\n<td style=\"padding: 0px 10px;min-width: 150px;border: none !important\">\n<p><strong>Gabriel Pedroza<\/strong> (<a href=\"https:\/\/www.linkedin.com\/in\/gabriel-pedroza-89bb5338\/\">LinkedIn<\/a>) is a Principal Research &amp; Development Engineer in Safety-Security at Synopsys. He specializes in the safety, certification, and trustworthiness of AI\/ML-based systems in critical domains, leveraging model-based engineering and formal methods to address robustness and compliance challenges.<\/p>\n<\/td>\n<\/tr>\n<\/table>\n","protected":false},"template":"","class_list":["post-198982","topic","type-topic","status-publish","hentry","topic-tag-ai","topic-tag-ai-ml","topic-tag-iso-pas-8800","topic-tag-ml","topic-tag-safety","topic-tag-scade","topic-tag-standards"],"aioseo_notices":[],"acf":[],"custom_fields":[{"0":{"_edit_lock":["1780601963:1769"],"_edit_last":["1769"],"_aioseo_title":[null],"_aioseo_description":[null],"_aioseo_keywords":["a:0:{}"],"_aioseo_og_title":[null],"_aioseo_og_description":[null],"_aioseo_og_article_section":[""],"_aioseo_og_article_tags":["a:0:{}"],"_aioseo_twitter_title":[null],"_aioseo_twitter_description":[null],"filter_by_optics_product":["Lumerical"],"_filter_by_optics_product":["field_64fb192ba3121"],"application_name":[""],"_application_name":["field_64a80903c8e15"],"family":[""],"_family":["field_64a809229a857"],"siebel_km_number":[""],"_siebel_km_number":["field_63ecbffce60db"],"salesforce_km_number":[""],"_salesforce_km_number":["field_63ecc018e60dc"],"km_published_date":[""],"_km_published_date":["field_64c77704499dd"],"product_version":[""],"_product_version":["field_64c776cb4fd2e"],"_bbp_forum_id":["27825"],"_bbp_topic_id":["198982"],"_bbp_author_ip":["87.190.19.71"],"_bbp_last_reply_id":["0"],"_bbp_last_active_id":["198983"],"_bbp_last_active_time":["2026-06-04 19:34:32"],"_bbp_reply_count":["0"],"_bbp_reply_count_hidden":["0"],"_bbp_voice_count":["0"],"_btv_view_count":["55"]},"test":"solution"}],"_links":{"self":[{"href":"https:\/\/innovationspace.ansys.com\/knowledge\/wp-json\/wp\/v2\/topics\/198982","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/innovationspace.ansys.com\/knowledge\/wp-json\/wp\/v2\/topics"}],"about":[{"href":"https:\/\/innovationspace.ansys.com\/knowledge\/wp-json\/wp\/v2\/types\/topic"}],"version-history":[{"count":1,"href":"https:\/\/innovationspace.ansys.com\/knowledge\/wp-json\/wp\/v2\/topics\/198982\/revisions"}],"predecessor-version":[{"id":198983,"href":"https:\/\/innovationspace.ansys.com\/knowledge\/wp-json\/wp\/v2\/topics\/198982\/revisions\/198983"}],"wp:attachment":[{"href":"https:\/\/innovationspace.ansys.com\/knowledge\/wp-json\/wp\/v2\/media?parent=198982"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}