Powered by IoT technologies, smart home systems effectively support aging-in-place. However, they have some limitations in comprehensive event perception and timely appropriate action. Homecare robot systems have been proven to be effective in homecare task executions but still have significant technological challenges. To address this problem, this paper proposes a smart home system architecture integrating a mobile robot with better event perception and task execution performance. To support the proposed system architecture, an ontology of the smart home is built to address the data heterogeneity issue. Then an event perception method is built upon multi-data integration. To perform complex tasks, this work implements an improved genetic algorithm for task planning. Finally, simulations and physical experiments are conducted to validate the feasibility of the proposed system architecture.
