An intelligent system for smart buildings using machine learning and semantic technologies: A hybrid data-knowledge approach

The Internet of Things allowed us to seamlessly integrate communication and computational capabilities into everyday things that resulted in a technologically enhanced environment. However, we still need to work on integrating high level understanding and intelligence in this connected system. The IoT is a mean that enables the possibility of integrating intelligent behavior and services into surrounding environments. One of the most representative examples of artificial environments are buildings. Residential buildings (e.g. homes, apartment blocks) or dedicated public buildings (educational, medical, commercial, governmental) serve different purposes and needs, and therefore they have different characteristics and constraints. However, every building uses some form of resource (e.g. energy, water) in order to assure the required level of comfort, safety and conditions for carrying out the desired activities. In this paper we take a look at some questions regarding the construction and the exploitation of knowledge related to different types of buildings in order to optimize the use of different resources while still assuring the occupants' comfort. We enumerate some of the elements that characterize a building as smart and finally, we present a model for a building management system based on hybrid knowledge.