Architecture and Features of a Mobile Cloud Testing Platform

In early 2011 the author discovered a market opportunity for mobile cloud testing, initiated a demo project, and later filed a patent for a cloud testing platform.

After joining JD Finance in 2017, a simple system supporting automation and client performance testing was built, leading to a successful proof‑of‑concept for a state‑owned enterprise and the launch of JD Finance's cloud testing platform.

The platform consists of four main parts: Record‑Replay (script recording, playback, and conversion to Appium/Uiautomator/Robolectric), Test Management (management of test assets such as scripts, data, cases, and results), Test Service Management (service‑oriented interfaces for various test types like automation, performance, compatibility, stability, and security), and Cloud Device Management (centralized control of Android and iOS devices via SDKs and WDA, enabling web‑based access to devices).

Key technical details include support for sequential, branch, and loop structures, parameterization, checkpoints, and script segmentation; a semantic rule set for translating scripts to mainstream tools; and the use of Uiautomator with minicap/minitouch for Android interaction.

Challenges identified are insufficient device inventory, network‑induced inefficiencies, stability issues with long‑term device usage, and compatibility overhead when onboarding new devices.

Thoughts and suggestions highlight the potential of AI‑driven image recognition to enhance record‑replay, the need for streamlined test case and data organization, and the consideration of private‑cloud deployments for medium‑sized teams to improve efficiency and stability.