Like me, you're probably hearing a lot about AI (Artificial Intelligence) in software test automation. All the software-houses are pushing it, all the consultancies are talking about it. It's one of the in-vogue phrases, like test automation once was, way back when.
It's easy to get caught up in the hype, to buy into what you're being sold. Providers are often prone to hyperbole, bigging things up more than they deserve. You can come away from a presentation expecting the latest test tools to run themselves. This is not the reality, this is not Hollywood, and they are not Cyberdyne Systems or HAL.
The fact is, AI is not a testing silver bullet, but it will make your life easier.
Test Automation has evolved
I remember a time, not too long ago, where people thought automation was the testing panacea. It could reduce cost, improve velocity, accelerate release frequency, and reduce overheads - whatever the industry buzzword, test automation could achieve it.
Test Automation has always had its place, but it wasn't easy to set up, at least early on. Test Automation has matured significantly over the years and is more robust than ever. Even now though, to be successful, traditional test automation needs the fundamentals in place
- Business requirements
- Prioritised processes
- Solid and predictable data
- And more...
You can't just drop automation into a project and expect an instant solution, and so it is with AI.
AI is in the early stages; pushing the limits of what's possible with automation, but it needs the right things in place. As I said, AI is not going to solve all your problems on its own, but it will make your life easier.
Test AI - Evolution, Not Revolution
AI is the next stage in the evolution of test automation. AI can take things to the next level if employed correctly. But, just like traditional test automation, AI needs the basics in place, and you need to know the limits of what it can do today. You still need to know what you're testing, what a pass looks like, what data to use etc. Once you have that set up though, AI brings benefits.
Test Automation AI reduces maintenance time.
In changing environments, object recognition often causes scripts to fall over. Take old versions of UFT as an example; without AI, objects are scanned and stored in the Object Repository (OR). The OR records all sorts of information about the object and allows UFT to know exactly which object to interact with. The OR is a great solution, until a developer or designer changes (even moves) an object. Sometimes even a minor change can knock your script over, forcing you to rescan the object.
Instead of building an OR, the latest version of Micro Focus UFT One uses AI to find an object by the smallest of details; its name for example. It doesn’t need to know anything else, no scanned list of details. This means that even if that object moves or other characteristics change, UFT can still find it. Your script doesn't fall over, your maintenance time is eliminated.
Granted, the AI script won't run as fast as an OR based script – searching for the objects takes a little longer – but it will be much more resilient.
It’s worth noting that this feature has existed for a while in test automation frameworks, such as SAP TAO, but you’re now getting it out of the box, without significant investment or setup. It’s proven and robust, but you still need to automate the test in the first place.
Test Automation AI Reduces Your Test Development Time
Reduced maintenance is a significant benefit. AI can also help you create test-packs, that is, select appropriate tests to run. AI can search for additional routes to pre-identified defects, giving your developers more context and allowing quicker fixes.
With Test Automation AI You Can Use The Same Script on Different Operating Systems
Using AI, Micro Focus UFT One allows you to execute a single test script across multiple operating systems. For example, if you need to test a mobile app on iOS and Android. Without AI, you need separate scripts for each operating system. With AI, a single script works across both.
AI Could Be The Differentiator You Need
We all need to be realistic about AI. We’re not at T-800 or HAL 9000 quite yet, and you don’t need to start looking for a new career. But AI is here and it’s already helping test teams get through their workloads…
- AI is already adding robustness to tests
- AI is already reducing the need for user interaction
- AI is already helping developers fix their solutions
As always with technological change, it’s a good idea to understand its potential as early as possible. Could you harness AI for your own benefit? Remember, it's included free in the leading test automation products.
You never know, test Automation AI could be the differentiator you need to drive super-slick testing through your test team.
Post a Comment
Sign up to receive the latest, software testing insights, news and to join the Calleo mailing list. You can, of course, unsubscribe at any time.