Proactive Fuzzy Control and Adaptation Methods for Smart Homes

Proactive, context-aware computing isn´t new. In 2000, David Tennenhouse called for a change in the boundary between the physical and virtual worlds. He identified proactive computing as an alternative to interactive computing and defined how future systems should become more involved with the real world. He also considered context-aware control systems with online adaptation especially promising. Today, ambient-intelligence researchers show increasing interest in both proactive applications and context-aware applications. Using different context-recognition methods, researchers can easily gather application-specific information from the environment and enable context-triggered actions. According to Hee Eon Byun and Keith Cheverst, a context-aware home can serve its inhabitants more flexibly and adaptively than an ordinary home. They also claim that proactive systems can be built using machine-learning algorithms with context recognition. In addition to context recognition, adaptivity is essential in intelligent environments. In terms of computing systems, the environment can adjust itself using adaptation mechanisms to comply with user preferences; it can become unobtrusive and better support user activities. A home adapting to its inhabitants´ living style is much more convenient than a user adapting to the home´s behavior. In this article, we show how to develop a proactive, adaptive, fuzzy home-control system, present the algorithm we used for adaptation, and evaluate the test results we obtained.