2018 |
Renoux, Jennifer; Klügl, Franziska Simulating Daily Activities in a Smart Home for Data Generation Inproceedings Proceedings of the Winter Simulation Conference, Gothenburg, Sweden, 2018. Abstract | Links | BibTeX | Tags: constraint-based planning, simulation, smart home @inproceedings{renoux2018simulating, title = {Simulating Daily Activities in a Smart Home for Data Generation}, author = {Jennifer Renoux and Franziska Klügl}, url = {http://www.jenniferrenoux.com/wp-content/uploads/2018/10/main.pdf}, year = {2018}, date = {2018-12-09}, booktitle = {Proceedings of the Winter Simulation Conference}, address = {Gothenburg, Sweden}, abstract = {Smart Homes are currently one of the hottest topics in the market of Internet of Things and Augmented Living. However, in order to provide high-level intelligent solutions, algorithms need to be developed that take into account which activities the inhabitants intend to perform. Sensor data plays an essential role in their development, ranging from testing the algorithms to learning underlying rules for classifying and connecting sensor patterns and to inhabitant activities. However, currently only few and limited data sets are available. In this contribution we present concepts and solutions for generating high-quality data using a flexible agent-based simulation tool. Hereby, we integrate the simulation of a sensorized apartment with human behaviour modeling. We selected a constraint-based planning approach for producing a sequence of daily activities of a human inhabitant. The overall set-up is shown to produce data that exhibits the same relevant properties than a comparable real-world scenario and thus can be used to replace expensive data collection campaigns.}, keywords = {constraint-based planning, simulation, smart home}, pubstate = {published}, tppubtype = {inproceedings} } Smart Homes are currently one of the hottest topics in the market of Internet of Things and Augmented Living. However, in order to provide high-level intelligent solutions, algorithms need to be developed that take into account which activities the inhabitants intend to perform. Sensor data plays an essential role in their development, ranging from testing the algorithms to learning underlying rules for classifying and connecting sensor patterns and to inhabitant activities. However, currently only few and limited data sets are available. In this contribution we present concepts and solutions for generating high-quality data using a flexible agent-based simulation tool. Hereby, we integrate the simulation of a sensorized apartment with human behaviour modeling. We selected a constraint-based planning approach for producing a sequence of daily activities of a human inhabitant. The overall set-up is shown to produce data that exhibits the same relevant properties than a comparable real-world scenario and thus can be used to replace expensive data collection campaigns. |
2018 |
Simulating Daily Activities in a Smart Home for Data Generation Inproceedings Proceedings of the Winter Simulation Conference, Gothenburg, Sweden, 2018. |