Professional Courses & Training

Professional Courses & Training

di Veljko Massimo Plavsic
Corsi

ISO/PAS 8800. Lesson 3:Relation with ISO 26262 and SOTIF

1. Overview of the Safety Ecosystem ISO/PAS 8800, titled 'Road vehicles — Safety and artificial intelligence', was developed to address the unique safety challenges posed by Machine Learning (ML) and Artificial Intelligence (AI) in automotive applications. It does not replace the existing safety standards; rather, it acts as a specialized supplement. To understand its role, one must look at the two primary pillars of automotive safety: ISO 26262 (Functional Safety) and ISO 21448 (Safety of the Intended Functionality, or SOTIF). 2. Interaction with ISO 26262 (Functional Safety) ISO 26262 focuses on hazards caused by malfunctions in electrical and electronic (E/E) systems. These are typically divided into systematic failures (e.g., software bugs) and random hardware failures. How ISO/PAS 8800 Fits: While ISO 26262 provides the general framework for software development (Part 6), it was not originally designed for the non-deterministic nature of AI. ISO/PAS 8800 provides specific guidance for the 'AI element' within the ISO 26262 lifecycle. It helps define how to handle systematic failures in the AI training process, model selection, and deployment that could lead to functional safety violations. ## 3. Interaction with ISO 21448 (SOTIF) SOTIF deals with hazards that occur without a system failure. Instead, these hazards arise from performance limitations or environmental triggers (e.g., a vision system failing to detect a pedestrian because of intense sun glare). How ISO/PAS 8800 Fits: AI performance limitations are a core concern of SOTIF. ISO/PAS 8800 expands on the SOTIF concept by providing detailed methodologies for AI-specific issues like data bias, over-fitting, and robustness against adversarial attacks. It provides the technical 'how-to' for achieving the safety goals defined by the SOTIF process when AI is the underlying technology. ## 4. The Integrated Approach The relationship can be visualized as a Venn diagram where ISO/PAS 8800 sits at the intersection of AI development and automotive safety requirements. ISO 26262: Ensures the AI hardware and integration logic don't break. ISO 21448 (SOTIF): Ensures the AI's intended function is safe in complex environments. ISO/PAS 8800: Provides the specific AI/ML engineering practices to satisfy both of the above. ## 5. Key Mapping Points Data Quality: ISO/PAS 8800 provides requirements for dataset completeness and representativeness, which supports SOTIF's goal of reducing 'Unknown Unsafe' scenarios. * Validation & Verification: It introduces AI-specific V&V methods, such as metamorphic testing, which are required to supplement the traditional testing methods found in ISO 26262.
10 apr 2026
1:20

ISO/PAS 8800 Lesson 2:Scope and Regulatory Context

Lesson: Scope and Regulatory Context of ISO/PAS 8800 1. Introduction to ISO/PAS 8800\nISO/PAS 8800, titled "Road vehicles — Safety and artificial intelligence," is a Publicly Available Specification designed to provide a dedicated framework for the safety-related aspects of Artificial Intelligence (AI) in automotive applications. As vehicles become increasingly automated, traditional functional safety standards like ISO 26262 reach their limits, particularly regarding the non-deterministic nature of machine learning (ML). ISO/PAS 8800 bridges this gap by offering guidance on how to integrate AI within the existing automotive safety ecosystem. 2.Scope of the Specification The scope of ISO/PAS 8800 is precisely defined to ensure it addresses the unique challenges of AI without duplicating existing standards. It focuses on: Machine Learning Life Cycle: From data collection and labeling to model training, verification, and deployment. Safety-Related AI Systems: It applies specifically to AI components that contribute to the safety-related functions of the vehicle (e.g., perception systems in ADAS or autonomous driving levels 3-5). Interplay with Existing Standards: ISO/PAS 8800 does not replace ISO 26262 (Functional Safety) or ISO 21448 (SOTIF). Instead, it provides the AI-specific methodologies needed to satisfy the requirements of those standards. Out of Scope Items Non-safety-related AI (e.g., personalized infotainment or comfort settings).General AI ethics (addressed by other standards like ISO/IEC 42001). Detailed hardware design (covered by ISO 26262). 3. Regulatory Context and the Global Landscape\nThe automotive industry operates under a complex web of regional and global regulations. ISO/PAS 8800 serves as a technical foundation that helps manufacturers demonstrate compliance with high-level legal requirements. The EU AI Act The European Union's AI Act classifies certain automotive AI applications as "high-risk." ISO/PAS 8800 provides a technical roadmap for meeting the Act's requirements regarding data governance, transparency, and robustness. UN Regulations (WP.29) UN Regulation No. 157 (ALKS) and the ongoing developments in the World Forum for Harmonization of Vehicle Regulations require rigorous safety proofs. ISO/PAS 8800 offers the standardized language and metrics needed to present these proofs to regulatory bodies. 4. Why ISO/PAS 8800 Matters Now Before this specification, manufacturers used disparate, proprietary methods to validate AI. This lack of standardization created uncertainty for regulators and consumers alike. ISO/PAS 8800 creates a shared baseline, ensuring that "safety" means the same thing whether the AI was developed in Munich, Silicon Valley, or Tokyo.
10 apr 2026
1:39

ISO/PAS 8800 Lesson 1: Introduction to ISO/PAS 8800 - Bridging AI Innovation and Automotive Safety

IA
Lesson 1: Introduction to ISO/PAS 8800 - Bridging AI Innovation and Automotive Safety 1.1 The Evolutionary Context: Why AI Needs a New Safety Paradigm For decades, automotive safety was governed by the principles of Functional Safety (ISO 26262), which focuses on hardware reliability and the mitigation of systematic software errors through rigorous, deterministic logic. However, the paradigm shift toward Artificial Intelligence (AI) and Machine Learning (ML) has introduced a level of complexity that traditional standards cannot adequately address. The Historian’s Perspective: From C-Code to Neural Weights Historically, vehicle safety was built on the premise of "code you can read." If a sensor detects an obstacle, a deterministic line of code triggers the brake. In the AI era, specifically with Deep Learning, logic is replaced by millions of parameters (weights) learned from data. The "historian" notes that while ISO 26262 is excellent at catching a bit-flip in memory or a software bug, it is not equipped to handle the probabilistic nature of a neural network that might misclassify a stop sign because of a specific shadow pattern. The Necessity of ISO/PAS 8800 ISO/PAS 8800 (Road vehicles — Safety and artificial intelligence) was published to address this specific "black box" challenge. It serves as the industry’s response to the realization that AI is not just another software module, but a fundamentally different way of processing information that requires a bespoke safety framework. 1.2 Scope: The AI Lifecycle under the Safety Lens ISO/PAS 8800 provides a comprehensive guide for managing the safety of AI-based systems throughout their entire lifecycle. Unlike traditional software development, which focuses on the "V-Model" of design and testing, AI safety focuses heavily on the Data Lifecycle and Model Robustness. Key Areas of Focus: Data Quality and Lineage: Ensuring that training data is representative, unbiased, and free from artifacts that could lead to unsafe behaviors. This includes the documentation of data sources and preprocessing steps. The Learning Process: Standardizing how models are trained, including the selection of loss functions and hyperparameter tuning to ensure predictable outcomes. Model Robustness and Generalization: Addressing how an AI performs when it encounters "Out-of-Distribution" (OOD) data—scenarios it did not see during training. Deployment and Monitoring: Establishing safety bounds for AI performance in real-time and determining when a human or a secondary system must intervene. 1.3 The Automotive Safety Trinity: Relationships and Synergies One of the most common misconceptions for beginners is that ISO/PAS 8800 replaces previous standards. In reality, it forms a critical third pillar in a unified safety strategy.
10 apr 2026
0:59

Lesson 2: Advanced Context Management and Role Definition

IA
Explore sophisticated techniques for defining complex personas, managing lengthy context windows, and maintaining consistency across multi-turn conversations.
27 mar 2026
1:53

Claude Fundamentals and Prompting Best Practices

IA
The Goal: What exactly should the output achieve? The Audience: Who is reading this output? The Format: Required structure (e.g., JSON, Markdown list, essay, dialogue). Example of Improvement: Poor: "Write about renewable energy." Strong: "Act as a sustainability consultant. Draft a 5-point executive summary, formatted as a numbered Markdown list, comparing the long-term CapEx viability of solar photovoltaic versus offshore wind energy for a European manufacturing client." B. Role Assignment (Persona Prompting) Assigning a clear, defined persona significantly sharpens the tone, vocabulary, and focus of the response. This leverages Claude's vast training data by forcing it to adopt a specific knowledge domain. Best Practice: Always start by stating the role: You are a Senior Data Scientist specializing in Python and Pandas. C. Instruction Placement and Delimiters For complex prompts, use delimiters (like triple quotes """, XML tags , or triple backticks ```) to clearly separate instructions from the input data or context Claude needs to process. This reduces the risk of the model confusing instructions with content. D. Few-Shot Learning (In-Context Examples) When the desired output style is highly nuanced or specific, providing one or more examples (Input/Output pairs) drastically improves adherence to the pattern. This is particularly effective for structured data transformation or stylistic imitation. E. Iterative Refinement Rarely is the first prompt perfect. Intermediate prompt engineering involves treating the interaction as a dialogue where you refine instructions based on the previous output. Use phrases like, "That was good, but now focus more on the financial implications," or "Reformat the third point to use active voice only." By mastering these fundamentals, you create a stable baseline, making the transition to advanced techniques like Chain-of-Thought (CoT) and structured data extraction much smoother.
27 mar 2026
2:06

Introduction to Radiometric Data and Measurement Tools

IA
High humidity or fog can scatter IR radiation. While often minor for short industrial inspections, it becomes critical for long-range outdoor work. 3. Using On-Camera Measurement Tools Your FLIR camera comes equipped with several standard tools to extract and analyze radiometric data directly in the field. We will focus on the primary tools: 3.1 Spot Meter (Point Measurement) This is the most fundamental tool. It places a single crosshair on the screen, displaying the temperature reading for the pixel directly underneath it. Use Case: Checking the temperature of a single bearing, bolt, or component. Configuration: Ensure the spot meter is referencing the correct measurement mode (e.g., Max, Min, or Average, if applicable). 3.2 Area Box (Box Measurement) This tool allows you to draw a rectangular box over a region of interest (ROI). The camera displays several statistics for all pixels within that box simultaneously: Max (Maximum): The single hottest temperature reading within the box. Min (Minimum): The single coldest temperature reading within the box. Average (Mean): The average temperature across all pixels in the box. Use Case: Assessing the overall temperature of a motor casing, insulation coverage, or identifying the hottest spot within a problematic area. 3.3 Line Measurement Some advanced models allow drawing a straight line across the image. The camera then plots a temperature profile graph along that line. Use Case: Analyzing temperature gradients across a surface, such as checking for even heat distribution across a heating element or insulation failure profile. 4. Saving and Interpreting Radiometric Files When you save an image on a radiometric camera, you are not just saving a JPEG. You are saving a file structure (often a JPEG overlaid with proprietary metadata or a dedicated radiometric format). This metadata block contains all the settings you configured: Emissivity, Reflected Temperature, Distance, Date/Time, and the calibration curve used by the detector. Why this matters: You can load this file later into FLIR's post-processing software (like FLIR Tools or ResearchIR) and change the emissivity or reflected temperature settings without losing the raw data, allowing you to recalculate accurate temperatures after leaving the site. This is the true power of radiometric thermal imaging.
27 mar 2026
2:05

Optimizing Image Quality through Settings

measuring small components or when the object is very close or very far. Always fine-tune the focus just before taking a measurement on a critical spot. --- Summary of Optimization Steps Set Emissivity (ε): Match the surface material value. Set Background Temperature: Apply reflectivity compensation if necessary. Select Palette: Use Iron or Rainbow for discovery; switch to Gray Scale for documentation. Adjust Span: Narrow the span around your expected temperature range for maximum contrast. Ensure Sharp Focus: Manually adjust if automatic focus fails or when inspecting small targets. By mastering these settings, you transition from simply seeing heat to quantitatively analyzing thermal performance
27 mar 2026
2:45

Understanding and Interpreting Thermal Images

IA
boundaries of a defect. 3.3.3 Reporting Standards Every finding must be reported clearly: Location: Precise identification (e.g., Panel A, Breaker 14, Phase C). Image: Include both the thermal image and the corresponding visible light image (if available). Data: Record the Max Temp, Background Temp, Emissivity used, and Ambient Temp. Classification: Assign a severity level (e.g., Minor Deviation, Action Required, Critical Fault) based on established standards (e.g., Infrared Training Center (ITC) guidelines). By mastering color interpretation, understanding the physics of emissivity, and rigorously documenting findings using analytical tools, you transform raw data into valuable diagnostic information.
27 mar 2026
2:20

Basic Camera Operation and Setup

IA
Excellent for sharp delineation of hot/cold areas. * Greyscale/White Hot: Best for documentation where visual realism is preferred. 2.4 Focusing and Image Capture Sharp images are critical for accurate analysis. Thermal images require focus, just like visible light cameras, though the process is different. Focusing Methods Automatic Focus (AF): Pressing the focus button once usually snaps the image into focus based on the center reading. Manual Focus (MF): If AF fails (common with very distant or very close objects), use the joystick or dedicated focus ring/buttons to manually adjust until the image appears sharpest. Capturing and Saving Images Framing: Frame the area of interest, ensuring the target fills a sufficient portion of the screen. Focus & Adjust: Ensure focus is sharp and contrast/level adjustments (span) are optimized. Capture: Press the shutter button completely. The camera will momentarily freeze the image and prompt you to save it. Saving: Select 'Save' on the screen or press the shutter button again (depending on the model configuration). Always add a brief note or inspection ID if the camera supports annotation features. 2.5 Reviewing and Managing Files To review previously captured images, exit the live view screen and enter the image gallery (usually indicated by a folder icon). Navigation: Use the directional controls to scroll through images. Image Analysis: When reviewing a saved image, the camera often allows you to move the spot meter, change the color palette, or adjust emissivity after capture (metadata adjustments), which is a powerful diagnostic tool.
27 mar 2026
2:25

FLIR camera hardware and components

close distance, ideal for scanning large surfaces quickly. Narrow-Angle Lenses (e.g., 13°): Provide a much smaller FOV, allowing you to measure smaller targets from much farther away. These lenses offer better Spatial Resolution (the ability to see small details). Spot Size Ratio (SSR) and Distance Measurement This is perhaps the most crucial concept related to the lens. The SSR defines the ratio between the distance to the target and the smallest spot size the detector can accurately measure.
27 mar 2026
2:14
3 di 5