Analysis and testing of notifications in Android wear applications

Abstract

Android Wear (AW) is Google's platform for developing applications for wearable devices. Our goal is to make a first step toward a foundation for analysis and testing of AW apps. We focus on a core feature of such apps: notifications issued by a handheld device (e.g., a smartphone) and displayed on a wearable device (e.g., a smartwatch). We first define a formal semantics of AW notifications in order to capture the core features and behavior of the notification mechanism. Next, we describe a constraint-based static analysis to build a model of this run-time behavior. We then use this model to develop a novel testing tool for AW apps. The tool contains a testing framework together with components to support AW-specific coverage criteria and to automate the generation of GUI events on the wearable. These contributions advance the state of the art in the increasingly important area of software for wearable devices.