There are several interesting web app automation scenarios that we can improve using AI:
- Increase test execution stability (self-healing automation) by letting AI to automatically locate web elements when the primary locators fail. This feature already appears in some cutting-edge automation tools like Mabl.
- Increase automation productivity by using Natural Language Processing (NLP) to automatically translate manual test cases into automated test cases. In theory, this capability should be feasible but I haven’t seen any open-source or commercial solutions available on the market yet.
- Increase test coverage by auto-generate input parameters that could exhaustively test the API under test using advanced algorithms such as Pairwise.
- Reduce test execution time by only running the tests that are impacted by a certain code commit instead of exhaustively running all of them after every code commit. For instance, check out the Test Impact Analysis feature of Microsoft Azure DevOps.
- Reduce failure debugging time by auto analyzing test results and assigning them to categories like Environment Issues, Automation Problems, App Defects, etc. An outstanding example of this category is Report Portal.
- Perform visual validations on web apps using OCR or image-recognition techniques. There are plenty of tools out there such as TestArchitect, AppliTools and SikuliX.
AI is just one trend in the industry. Read more about other trends and predictions here: 2019 Test Automation State-of-the-Practice and Trends | LogiGear Magazine