AI and data-driven methods, particularly machine and deep learning models, have revolutionized image processing tasks such as object detection and segmentation, as well as inverse sensing tasks such as atmospheric condition monitoring. These models enable highly accurate predictions from large, labelled datasets. However, one of the key challenges with current AI systems is their opacity. While they deliver precise results, they often fail to provide clear insights into the underlying mechanisms driving these predictions. The lack of interpretability, especially in the context of sensing data, complicates the understanding of how internal features are represented and utilized. This issue is particularly pronounced in smart environment applications, where artificial intelligence techniques are increasingly employed, yet their internal workings remain largely opaque.

• Introduction
• Chapter 1: Adversarial machine learning attacks in IoT networks
• Chapter 2: Advanced security mechanisms for heterogeneous sensing data in IoT ecosystems
• Chapter 3: AI-enabled sensing, security, and explainability for smart urban, industrial, and environmental applications
• Chapter 4: AI-powered environmental risk assessment leveraging multi-sensor fusion
• Chapter 5: Synthetic data strategies for indoor human activity recognition: analytical models and sensor-based measurement campaigns
• Chapter 6: Intelligent resource allocation for 6G-enabled IoT sensing networks in smart cities
• Chapter 7: Trustworthy resource allocation in AI-driven 6G sensing networks
• Chapter 8: AI-enabled Wi-Fi and RFID sensing for smart environments: enhancing trust, privacy, and spatial resolution
• Chapter 9: Adversarial machine learning countermeasures in IoT networks
• Chapter 10: AI and Blockchain applications in IoT security systems: fundamentals, challenges, and opportunities
• Conclusion