Software Engineering is yet to find one solution where the probability of success in meeting all objectives is high. Even though there is much research happening, people are still believing in the process of fine-tuning as one of the solutions. The current trend is to adopt an agile methodology with a shorter sprint cycle of delivery to get customer feedback faster. Tool penetration in the Software Development Life Cycle also reduced human errors, improved productivity, and reduced variance significantly. This is one engineering field where tools are still evolving in each phase. Success in one phase is making people control schedule, quality, and cost variance in projects significantly. That phase is the functional testing phase.
In this blog, I would like to share some of the challenges in functional test automation and the approach which worked for me while consulting organizations.
I summarize the challenges I heard which you may hear in some of your current business situations.
- Number of planned releases to the system testing team and actual releases have a huge variance
- There is a feeling that certain issues in software should have been caught earlier in the life cycle
- Test cases are in place why are we not automating test execution? Which framework to adapt? Which tool to use?
- Our solution is complicated we have many environments on which we need to test and there is no single framework that helps Web UI testing, API testing, Database Testing, and Mobility testing
- Why are we not using the same testing team who are testing currently, contributing to test automation effort also?
- We did automation using some framework but changes to the product were so much we could not sustain script maintenance speed resulting in scrapping many of the scripts which were working earlier
- Metrics we collect look too much but we are unable to answer some of the stakeholder's questions during the project review
So the search continues for one technical solution which will address many of them. I was also looking for such a solution to recommend to my client and I found it in a framework created by young engineers who had captured all challenges of test automation project failures very nicely and had addressed all of them in their framework. The people associated have used this to create young highly productive engineers who are 3 times faster in converting test cases into automated scripts, and have the capability of doing POCthein the shortest time due to 90% of the object handling required libraries they have in their framework. The same framework supports API testing, Database testing, and mobile side both Android and iOS applications.
The automation should support continuous integration and ensure high-quality release to the system testing team and later to production. Regression automation % also continuously should grow to enable faster cycle time to production release with reduced defect escapes. The percentage of projects not meeting the set objectives must reduce drastically to zero.