Artificial Intelligence and Machine Learning for Safety-Critical Systems

Artificial Intelligence and Machine Learning for Safety-Critical Systems: A Comprehensive Guide provides engineers and system designers who are exploring the application of AI/ML methods for safety-critical systems with a dedicated resource capturing the challenges and mitigation strategies involved in designing such systems. Divided into nine sections, the book covers the most important applications of safety-critical systems, helping readers understand how related problems are being solved in different domains/problem settings. The goal of this book is to help ensure that AI-based critical systems better utilize resources, avoid failures, and increase system safety and public safety. The authors present ML techniques in safety-critical systems across multiple domains, including pattern recognition, image processing, edge computing, Internet of Things (IoT), encryption, hardware accelerators, and many others. These applications help readers understand the many challenges that need to be addressed in order to increase the deployment of ML models in critical systems. In addition, the book shows how to improve public trust in ML systems by providing explainable model outputs rather than treating the system as a black box for which the outputs are difficult to explain. Finally, the authors demonstrate how to meet legal certification and regulatory requirements for the appropriate ML models.

• Covers foundational concepts, advanced theories, and real-world applications, ensuring readers gain a thorough understanding of AI/ML as it applies to Safety-Critical Systems
• Presents both the risks and advantages of implementing machine learning techniques in Safety-Critical Systems
• Presents machine learning techniques in Safety-Critical Systems across domains, including pattern recognition, image processing, edge computing, Internet of Things (IoT), encryption, hardware accelerators, and many others
• Demonstrates how to meet legal certification and regulatory requirements for the appropriate ML models