Over the last years, Ambient Intelligence (AmI) has been pointed out as an alternative to current practices in home care. AmI supports the concept of Ambient Assisted Living, which aims to allow older people to remain independent at their own homes for longer. The integration of a mobile robot into a database-centric platform for Ambient Assisted Living is described in this thesis. The robot serves as a rst-aid agent to respond to emergencies, such as a fall, detected by the intelligent environment. To accomplish that the robot must 1) be able to receive tasks from intelligent environment; 2) execute the task; 3) report the progress and the result of the task back to the intelligent environment.
The system of the robot is built on top of the Robot Operating System, while the existing intelligent environment on a PostgreSQL database. To receive tasks from the intelligent
environment, the robot maintains an active connection with the database and subscribes to specic tasks. A task, for example, is to nd a person in the environment, which includes asking if the person is doing well. To nd a person a map-based approach and a face reco ... mehrgnition are used. The robot can interact with people in the environment using text-to-speech and speech recognition. The active connection with the database enables the robot to report back about the execution of a task and to receive new or abort tasks. As a conclusion, together with an AAL system, mobile robots can support people living alone. The system has been implemented and successfully tested at Halmstad University on a Turtlebot 2. The code is available on Github.