Developing SPIM-TA: A Maturity-Level Framework for Systematic Process Improvement in Software Testing Automation

Abstract

The growing complexity and demands for high-quality software underlined the need for robust automated software testing frameworks. However, lack of skilled personnel, cost of testing automation, and maintenance of script along with others oppose the widespread adoption of automation in testing. The work proposed by this study is in line with the development of Systematic Process Improvement Model for Testing Automation, abbreviated as SPIM-TA. The model is inspired from Capability Maturity Model Integration (CMMI) and Software Outsourcing Vendor Readiness Model (SOVRM) maturity-level-based framework. The research was done based on a four-phased methodology. In the first phase, a systematic literature review identified 14 critical challenges in software testing automation with practices or solutions for each of these challenges. In the second round, a survey questionnaire is used to validate these findings in industrial context and extract more practical insights from case studies. This dual approach had ensured both academic rigor and the relevance of the practices with industry. The third one involved developing SPIM-TA by incorporating insights that come from the SLR and those results of the survey as well. The framework defines five maturity levels, systematically addressing the identified challenges and providing a structured pathway for organizations to enhance their automation processes. In the fourth phase, case studies of industrial applications were carried out to evaluate SPIM-TA's efficiency. The applicability, scalability, and impact of the framework on testing automation in the real world were assessed using the Motorola Assessment Tool. The findings illustrate the systematic way SPIM-TA will help overcome the automation challenge, so that organizations can reach more advanced levels of maturity and efficiency in their testing processes. In the first place, the present study closes the gap between academia and industry but at the same time paves the foundation for future research into the automation of software testing. © 2025 The Author(s)