Enhancing software test automation tools through Machine Learning and AI strategies : A case study of IT industry in Sri Lanka

Sammanfattning

This study aims to investigate the challenges of using current software automation tools and propose suggestions for using Artificial Intelligence (AI) and Machine Learning (ML). Employing a qualitative survey-based design, the study utilizes the Goal-Question-Metric (GQM) model to systematically translate the research hypotheses into directed content analysis. Data was collected via a structured questionnaire from 22 participants, selected through purposive sampling working in Sri Lanka as software automation testers. This study demonstrates that automating the testing process by using AI and Machine Learning can significantly improve the performance of software testing by enabling an automatic analysis of test data collected at different stages of testing. Key findings highlight that test script maintenance, dynamic element handling, scalability, and limitations of current automation tools are the main issues. Furthermore, the study identifies and discusses the most helpful AI/ML frameworks, which are Selenium, Applitools, Testim, MaBL, and IBM Watson AIOps. Based on our findings through collected data and literature review, these frameworks improve test case generation, defect detection and possess self-healing capabilities. Finally, the study proposes strategies for the integration of AI/ML into test automation in Sri Lanka. Technical training, investment and improved partnership with industry expertise are the main strategies.