International automated testing market may reach the $70 billion mark by 2032, according to some forecasts. It’s about 4 times higher than now. How realistic are such forecasts, what basis do they have? How promising is it now to develop the field of automated testing, and is there a possibility that as AI develops, the need for professionals in the field of automated testing will disappear?
We discussed the evolution of autotesting and the role of artificial intelligence with Anatolii Tymoshchuk, a recognized expert in the field of testing. Anatolii has in-depth knowledge in automated test platform development and significant experience working on large international projects.
During his professional career, Anatolii worked for Grid Dynamics, EPAM Systems, Andersen Lab and others. For example, he participated in the development of automated test solutions for the OPEN Cloud platform using Java, Selenium and Selenide. His contribution helped to improve the quality of services provided and increase the reliability of the company's cloud solutions.
Anatolii also participates in the training of young professionals, conducting lectures, seminars, and master classes.
Can you briefly remind us of the history of auto testing? How did it all begin?
Automation testing emerged as a natural evolution in software development to reduce manual testing time and improve product quality. In the 1950s, during the early stages of programming, tests were executed manually, with developers taking full responsibility. Later, as programming languages and technology evolved, the first scripts for verifying basic functionality were introduced.
In the 1980s-90s, with the rise of personal computers and complex systems, tools for recording and replaying test scenarios became popular, such as Mercury Interactive WinRunner. By the 2000s, open-source tools like Selenium (released in 2004) significantly expanded automation possibilities, especially for web applications. This marked the rise of various frameworks like JUnit, TestNG, and RestAssured for API testing.
Today, automation is an integral part of the software development process, covering UI, API, mobile applications, and even performance and security testing. The integration of automated tests into CI/CD pipelines allows quick defect detection and resolution, enabling more efficient development workflows.
Anatolii, as an experienced tester, you have worked with a wide variety of systems. What do you think has changed most in programs and processes in recent years?
In recent years, the most significant changes in programs and processes have been driven by automation, cloud technologies, and DevOps integration. Here are some key points:
1) Speed of development: Agile and DevOps have become standard in many companies. This has led to testing being integrated earlier in the development cycle, such as through Shift-Left Testing.
2) Tool advancements: More powerful automation tools have emerged, such as Playwright, Selenium for UI, and frameworks for API testing (RestAssured). They enable more stable and faster tests. 3) Cloud technologies:
Many applications are now hosted in the cloud, impacting testing by requiring consideration of dynamic infrastructure and leveraging cloud platforms for testing (AWS, Azure, LambdaTest). 4) AI and ML integration: Artificial intelligence and machine learning are increasingly used in testing processes for test generation, defect prediction, and result analysis. 5) Focus on security and performance: Due to the rise in cyberattacks, there is more emphasis on security testing. Performance testing requirements have also increased due to the scalability of modern systems. 6) Mobile app automation: Tools like Appium have become more integrated and efficient, enhancing mobile testing.
Overall, testing has become more integrated, automated, and focused on the rapid delivery of high-quality products.
What testing tools do you consider the best, and which ones are the worst? And why?
As regards the best tools, from my point of view, I could figure out 5 ones:
1) Selenium – a classic for web automation. Powerful and flexible, it supports many programming languages. Although it's aging, it remains a standard.
2) Playwright – a modern tool for web app automation. Faster and more stable than Selenium, it supports browsers, mobile devices, and API testing in one framework.
3) Selenide – an excellent extension of Selenium for Java. It simplifies test writing with built-in waiting methods and straightforward syntax.
4) RestAssured/Postman – perfect for API testing. RestAssured integrates well with Java projects, while Postman is convenient for quick requests and manual testing.
5) Jenkins – for CI/CD. Easy to set up for running automated tests and generating reports.
As regards less successful and less convenient tools, I could name
1) Katalon Studio – while easy to use, it’s too limited for large projects. The free version has restrictions, and the paid one is overly expensive.
2) QTP/UFT – expensive and less flexible for modern needs. It lags behind free alternatives.
3) Ranorex – cumbersome to use and not very efficient for large projects with frequent changes.
The best tools offer flexibility, ease of integration, and strong community support. They enable scalable, fast, and stable tests. The worst tools are typically outdated, have limited functionality, or are too expensive for what they offer.
What do you think about the rapid development of AI technologies, how could it affect the field of automated testing? And do you think there is a possibility that at a certain point the need for automated testing will simply disappear?
Test generation: AI can automatically create tests based on code analysis, specifications, or even user behavior history. This significantly speeds up test creation.
Result analysis: AI helps analyze large volumes of logs and test results, identifying defects and even predicting potential risks.
Self-healing tests: AI can identify and suggest fixes for tests that break due to changes in the code.
The need for automated testing won’t disappear, but its role will evolve. AI can automate routine tasks but won’t replace the expertise of testers in building strategies, designing complex tests, and defining requirements. The human factor remains critical, especially for testing complex business logic, UX/UI, and edge cases.
Furthermore, developing and maintaining AI systems will also require testing, opening up new opportunities for automation.
My conclusion: AI will become a powerful tool for QA engineers, but the need for testers will remain, shifting towards more intellectual and strategic work.
