LivingCare - An autonomously learning, human centered home automation system: Collection and preliminary analysis of a large dataset of real living situations

Research output: Contributions to collected editions/worksContributions to collected editions/anthologiesResearchpeer-review

Authors

Within the scope of LivingCare, a BMBF funded research project, a real senior residence was equipped with a large amount of home automation sensors. More than sixty sensors and actuators were installed in this apartment. This automa-tion system is working totally passive in the background. It doesn't perform any actions. All actions performed by hu-mans like switching light on or off, setting the temperature and the usage of electric devices like TVs will only be recorded as data and not performed by the system. This data is collected over a period of 18 months. Thus, one of the largest mobility and characteristics datasets based on home automation sensors will be acquired. This data will be the foundation for developing autonomously learning algorithms. During the second project phase these algorithms will start to control functions of the home automation system. The projects objective is to develop an autonomously learning home automation system that automatically adapts to the resident's behavior. The system will be able to grow with the users needs. With all the possible data collected it will be able to support daily actions, recognize behavior changes over timer and will be able to call help in emergency situations.
Original languageEnglish
Title of host publicationAmbient Assisted Living : 9. AAL-Kongress, Frankfurt/M, Germany, April 20 - 21, 2016
EditorsReiner Wichert, Beate Mand
Number of pages18
Place of PublicationCham
PublisherSpringer International Publishing AG
Publication date2017
Pages55-72
ISBN (print)978-3-319-52321-7
ISBN (electronic)978-3-319-52322-4
DOIs
Publication statusPublished - 2017

    Research areas

  • Informatics - home automation, Internet of Things, IoT, reinforcement learning, autonomously learning, real life data, field trail, Gait speed, Test Person, Sensor Event, Movement Sensor, Occupation Time