Semantic analysis and verification of context-driven adaptive applications in intelligent environments

With the advent of the Internet of Things (IoT) and Cyber-Physical Systems (CPS), assistive applications and smart environments can tap into a growing amount of context information to learn about their users, their surroundings and typical behavior. This information is useful to adapt intelligently, autonomously and non-intrusively to a myriad of circumstances. However, trustworthiness and reliable adaptation to changes in line with user expectations—especially to situations that the developers did not anticipate—remain key concerns. Understanding the impact of changes from a developer (design time) and system (runtime) perspective, and ensuring that no undesired side effects take place are two non-trivial research challenges to increase the adoption of such applications. Given the limited tool support for anticipating change at design time and runtime, we present our change impact analysis (CIA) methodology—found in the formal semantic modeling of intelligent environments and rule-based application behavior—to contribute to the development and deployment of reliable context-aware adaptive applications. We validate our contributions on non-trivial smart home and office scenarios, and demonstrate how our framework helps increase trust in intelligent environment applications by anticipating change implications upfront at design time and by minimizing the occurrence of undesired side effects at runtime. © 2016, Springer International Publishing Switzerland.