| || |
Automation project help
This is my first post and new to the forum.
At the moment I'm doing an internship in Software QA of a web interface belonging to a well known company.
While I have been here I have been exposed to some automation, including technologies/tools such as Selenium and TestNG.
I'm soon returning to finish my final year of college, in my final year we are required to do a big project. We are recommended to do something new such as developing a new tool for instance.
I'm hoping to develop something centered around automation, that would provide useful.
I have one idea that I'm currently messing around with, but not sure of going fully ahead with...
My question is , what is something that you as experienced individuals who are familiar with automation think there is lacking out there? What is something that you would think would be useful to you?
I'm not trying to rob ideas, or to get people to give me a project. I just wan't to have an idea to see what I could do and what I could think up of myself from your feedback. I don't want to go down the route of starting a project and realizing then its not going to be of use to anyone, or beneficial in anyway.
I think that's cool you're reaching out to the community for ideas to research/explore.
I can tell you things I have struggles with as someone who tries to bring automation into companies who have been doing manual for a long time.
* Total Cost of Ownership (TCO) / ROI calculations visualizations. It be great to generate using a simple webform an infographic of savings and ROI. I find the one of the most frustrating things about my job is people hire me with high aspirations to be 100% automated, but then they get paralyzed on where to start and get sticker shock on the costs, without realizing that they are paying way more in salaries and lost opportunities. Generally someone that takes me $2k in budget and 2 weeks of my time ends up with me spending 5 weeks putting together a presentation to convince them. It's the most frustrating thing, people hire me at and Sr. Engineer's salary then scrutinize over a few thousand.
* Measuring Automated Test Effectiveness. Automated tests are harder to maintain than production code. I want to be able to easily run reports on which tests pass/fail the most and how often they catch issues, and the severity of the issues they catch. Currently right now, I try to run all my test results into a TCM that pass/fail, but most TCM systems don't support associating test cases to automated test identifiers. I can run reports on how often a test case fails, but because automated tests and manual tests are not necessarily 1-to-1, it's hard to optimize which automated tests should be deprioritized or deprecated. (I tend not to run 1-to-1 automated to manual tests as manual tests tend to be written to too high of a detail that an automated test can validate more at the same time.)
* More on the technical side. A machine for running tests in the background based on what code the developer is editing. There's a cool tool called NCrunch that runs unit tests based on what files developers edit in a background process and let developers know the moment they break a unit test. It would be great for end to end tests.
* Refactoring tool for BDD. BDD/Gherkin has been a popular way of writing requirement specs as tests. However, they are annoying to refactor. I can't use tools such as ReSharper and intelligent editors to refactor large swaths of tests written in BDD. (Same is true for keyword tests)
(I'll post more as I think of them)
Last edited by dlai; 08-16-2016 at 03:33 PM.