Abstract: In order to make full use of multi-domain knowledge and improve the quality of service in the home environment, a multi-domain knowledge sharing and reuse method is proposed based on the home intelligent space. Firstly, an ontology-based multi-domain knowledge model is built in the light of the prior knowledge. On this basis, the software agent is used to achieve knowledge representation automatically, addressing the problem of data heterogeneity and facilitating the knowledge sharing and reuse. Then, according to the user's activities of daily living, an activity understanding approach is presented based on multiple sensors and an activity template to realize the user intention recognition and the service inference. Moreover, the acquisition of service sequences and the guidance of actions can be achieved by reusing the domain knowledge of human activity, and thus the calculation costs are reduced and the quality of service is improved. And the robot services are effectively executed through integrating the service execution strategy generated by the simulation system and the shared object knowledge in object domain, so as to improve the intelligence level of service execution. Finally, the experimental results are provided to show the effectiveness and feasibility of the proposed method.