This present study mainly describes the implementation and ideas of the mobile automation framework, which supports iOS and Android mobile automation testing technology. This study mainly uses a combination of qualitative and documentary analysis methods, designs some technical architecture diagrams, and writes some open-source code to implement respectively. Meanwhile, The automation code is not made public due to mobile automation security and privacy concerns. The paper is mainly driven by the so file inside Android and the ADB command, while iOS uses some class libraries based on the iOS system. The research framework also uses the Appium framework for encapsulation and research and carries out secondary encapsulation and call based on the internal Appium framework. It is convenient for the internal quality team's automated testing staff to use and execute and improve the efficiency and speed of automated software testing, which is the contribution of this research. Beside, The limitation of this research is that the framework is based on the secondary development and encapsulation of Appium framework, so this study are inevitably some imperfections and bugs that need to be continuously improved and used. Moreover, these contribution results demonstrate that it can combine the businesses of different china internet enterprises to complete. Next, the paper discusses specific cases of mobile automation testing framework and artificial intelligence in mobile framework design and evaluates their effects and impacts. Finally, the paper summarizes the roles and challenges of mobile automation testing framework design and AI enhancement elements and looks ahead to the future.