Curated Physical AI Insights
This library brings together SRES insights on the standards, engineering practices, and emerging developments that shape secure, responsible, and safe Physical AI development.
Browse the sections below to explore the topics most relevant to your work and see how they connect across a Physical AI safety framework. For all SRES insights, visit our Insights page.
Quick Navigation:
Standards Deep Dives
Engineering Methods and Practical Application
Emerging Standards and Industry Updates
Expert Perspectives and Fireside Chats
This library is reviewed and updated by SRES monthly.
Standards Deep Dives
These resources examine the standards, concepts, and relationships that shape functional safety and AI safety for Physical AI systems.
Safety Standards for Humanoid and General-Purpose Robots: A Practical Guide
A practical guide to the standards, regulatory considerations, and assurance approaches relevant to humanoid and general-purpose robots. It explains how intended use, mobility, manipulation, human access, AI-enabled behavior, operating environment, and target market shape the applicable framework, with four example safety pathways showing how these elements can fit together.
Comparing ISO/IEC TR 5469 and ISO/IEC TS 22440
A comparison of ISO/IEC TR 5469’s foundational guidance and ISO/IEC TS 22440’s emerging engineering requirements for AI safety-related control systems. It explains the shift from shared terminology and context toward actionable lifecycle activities, analysis, mitigation, and validation expectations.
Digging into ISO/IEC 5469:2024 – Artificial Intelligence: Functional Safety and AI Systems
An in-depth look at ISO/IEC TR 5469, a cross-industry reference for functional safety and AI systems. It examines the properties and risks introduced by AI and the ways AI can be used within, alongside, or to develop safety-related functions.
Engineering Methods and Practical Application
Standards establish expectations; engineering methods turn them into requirements, evidence, validation activities, and defensible safety arguments.
An examination of certification, audit expectations, and the broader applicability of ISO/PAS 8800 beyond automotive systems. It explains how AI development processes, traceability, dataset and validation evidence, and structured work products contribute to a defensible safety argument.
Emerging Standards and Industry Updates
Physical AI standards, regulatory requirements, and industry expectations continue to develop alongside a rapidly evolving technology landscape. These resources provide context for teams following those changes.
ISO/IEC TS 22440: The Emerging Direction for Functional Safety and AI Systems
An overview of the emerging ISO/IEC TS 22440 series and its role in addressing AI within safety-related systems. It explains how the series builds on IEC 61508 and other domain-specific functional-safety standards while incorporating concepts such as functional insufficiencies and triggering conditions.
Humanoid Robots in the Cybersecurity Era: The Impact of the EU Cyber Resilience Act
An overview of how the EU Cyber Resilience Act applies to connected humanoid and industrial robots, including product scope, key compliance deadlines, and cybersecurity obligations manufacturers should consider when developing products for the European market.
Expert Perspectives and Fireside Chats
This section brings together SRES perspectives and technical conversations on evolving Physical AI safety challenges. The fireside chats feature SRES partners Jody Nelson, Gokul Krithivasan, and Bill Taylor alongside special guests.
SRES Fireside Chat: Validation and Data in AI-Driven Systems
A conversation with SRES and Visage Technologies on verification, validation, simulation, synthetic data, data coverage, and assurance for AI-driven safety-critical systems. The discussion is relevant to automotive, autonomous, robotic, and other Physical AI applications.
Beyond Standards: Why Humanoid Robot Safety Requires What No Framework Has Codified Yet
A perspective on why traditional functional-safety approaches alone are insufficient for humanoid robots operating in dynamic human environments. It examines behavioral safety, SOTIF-style thinking, scenario-based validation, and the need to address risks that standards have not yet fully codified.
Humanoid Robot Safety Comes Into Focus
An early perspective on the safety challenges of humanoid robots operating close to people, including controlled shutdown, stability, scenario-based verification and validation, and the need for safety approaches that extend beyond traditional industrial robot safeguards.
Continue Exploring
This page highlights selected Physical AI resources rather than every SRES Insight. For the complete archive, explore SRES Insights or browse relevant topics:
Robotics & Physical AI · Functional Safety · Responsible AI · Cybersecurity · Autonomous Systems · Videos
For support applying these concepts to a specific product, organization, or development program, explore SRES Physical AI Consulting, including Robotics Safety, or Physical AI Training.
Join our family of safe and secure customers
