About the project
 

Most of the technologies developed for domestic applications have a basic goal: to simplify and make more comfortable the users lifes. However, these users should not be forced to adjust their habits due to new technologies. A litlle help these tecnologies can provide if  their benefits do not outweigh the difficulty of users adaptation. Thinking of a home automation system, the less the inhabitant of the residence has to know about technological aspects to interact with it, the more desirable the system will become.

Day-by-day, residents repeated actions that could become automatic. This project aims to develop a system that can automatically identify aspects of the environment and a specific moment that made a  user to take certain actions by repetition. This learning process is automatic and based on the behavior of people, avoiding the need to use an expert to design the entire system. Moreover, since habits of people change over time, i.e., an action took in appropriate time can no longer be taken on a future occasion, the system must be fully dynamic and reactive, changing its rules automation at all times.

therefore, The system learns the behavior patterns of the inhabitant and automates them over time. If a certain pattern is no longer used, the system will ignore the rules stablish for this specific pattern and creates new rules for new patterns that can be arisen. All this is possible by using some artificial intelligence algorithms known as inductive learning.

In this stage of the project, we have an automated system that can learn rules from a single inhabitant. We are now reseraching a method to detect and define specific rules for different inhabitants in a same environment.
Thus, a system of effective automation, as well as automatic learning, must have the ability to identify people in order to leran the suitable actions for each inhabitant.

For the identification system be considered effective, it should not require changing habits of the people in order to make the  identification possible. This projetc encompasses a research of a system that performs a inhabitant identification without his direct interaction with any type of biometric reader or camera. For this we propose the use of RFID, which is a technology developed to enable the identification of micro-electronic devices at a distance through radio frequency, in all clothes ans acessories of the inhabitants. The use of a group of RFID labels in clothes and accessories can allow a good identification of the inhabitant without ask him to change his way of life.

Thus, the system proposes a fully automated home automation, based on artificial intelligence techniques, which identifies people through combinations of clothes and accessories, and learns the habits of every person, providing convenience without requiring any adjustment in the way of life.