Integration of properties of virtual reality, artificial neural networks, and artificial intelligence in the automation of software tests: A review

Abstract

Abstract The complete automation of software tests has been considered to be an unattainable goal. This article discusses the potential to achieve this goal with recent discoveries and innovations in the areas of virtual reality (VR), artificial neural networks (ANNs), and artificial intelligence (AI). In this study, a theoretical proposal is described to integrate the properties of each of these areas using a process of automation of software tests. This process is based on a classification and description of the properties after consulting the literature, interviews, and dialogs with specialists from Australia, the United States, Germany, and Colombia. In addition to the experiences of the researchers, the construction of two tools is proposed: (1) a robot to design and apply functional tests, and (2) a virtual machine to identify errors in the logical structure of the code. Both tools are expected to replace human factors; the advantage is that the first tool identifies procedural flaws and the second errors of operation.