Automating Android Rotation Vector Testing in Google's Compatibility Test Suite Using a Robotic Arm
Resumo
Software testing is one of the essential phases of software development. In this context, test automation has gained significant traction for meeting immediate requirements while upholding result quality. While both automated and manual methods are employed for software testing, manual approaches are susceptible to inaccuracies and human errors despite the benefits of automation. This research is focused on Google’s Android OS, pivotal for a variety of global mobile devices, which must adhere to specific quality requirements. Our study focuses on the Compatibility Test Suite tool, specifically the Rotation Vector Computer Vision Crosscheck (RV) Test case designed for evaluating smartphone sensor capabilities. We introduce a three-axis robotic arm designed to automate the RV test execution, thereby minimizing operational failures and expediting Android smartphones’ quality testing processes. We compared RV test execution in a real-world company using our automated solution against human testers. The proposed robotic arm demonstrated a 75% accuracy rate, surpassing the 37% accuracy achieved by human testers. This significant disparity underscores the potential of our automation approach to mitigate manual errors while ensuring robust and effective testing processes.
