Why the telecom business needs scriptless testing with AI
Telecoms companies are under ever-increasing pressure to release quickly. One of the biggest obstacles, however, is testing.
How does testing keep up with the demand for secure, bug-free software, especially with the proliferation of devices? Automated continuous testing has addressed some of the challenges, but also created new ones.
Now, recent innovations in scriptless technology and AI-enhancements are being utilized by many industries to improve time-to-market, reduce costs, and productivity. For these reasons, scriptless testing with AI is expected to be one of the biggest trends in testing over the next couple of years. It offers many opportunities to the telecom industry.
Business benefits
By adding scriptless technology and AI to testing, organizations are gaining significant coverage in weeks, compared to many months of work. Lag time to better coverage and continuous testing shrinks and, importantly, is typically achieved without additional resources.
Plus, compared to traditional test automation methods — which many organizations have found hard to maintain and sometimes unreliable — scriptless testing removes flakiness of automation. AI and machine learning can make automation “self-healing” and reduce the maintenance overheads of tests and increasing their reliability.
As its name suggests, scriptless testing removes the need to code test scripts. Using AI and machine learning to enhance test automation, it has several significant benefits. For instance, test engineers (who are still responsible for most tests today) can use it to augment their work. Their time can be spent on more complex or higher priority testing and scriptless AI-based testing handling the rest.
However, one of the most potent advantages of scriptless, AI-based testing is that it makes it easier for anyone to carry out tests, not just qualified QA managers and test engineers. Business testers can reduce the time spent on manual tests and the risk of errors without needing specialist skills.
Telecom software developers can conduct functional testing more rapidly and take on responsibility for more tests – such as API tests – with minimal impact on their working days.
Moreover, AI-based testing offers built-in integrations to the test executions environment, browsers or devices, Continuous Integration (CI) pipeline and bug reporting. It also offers visual reporting, logs, various integrations to tools such as bug reporting.
If a telecom company is using open-source tools, it can develop these integrations itself. However, the time spent on that would be significant, while scriptless AI-based tools already offer these capabilities out of the box.
Implementing scriptless AI-based testing
While it has substantial benefits, it is crucial to bear in mind that scriptless AI-based testing represents a significant change to existing test strategies. There is a learning curve and impact on the organization’s culture to consider.
This is why the recommendation is to introduce scriptless AI-based testing progressively by fitting into the existing environment and starting small rather than disrupting the status quo on a large scale (risking failure and losing confidence that it is the right direction).
One tactic is to start with a handful of the most time-consuming and repetitive manual test cases in every regression and once they appear to be working, merge them into the CI/CD toolchain. If everything is still working and the tests having demonstrable value, then add another handful of tests, and so on.
Cultural communication
From a cultural perspective, it is vital to communicate the benefits of AI-based testing to everyone involved. Anything with an AI label is seen by some team members as a threat to their careers. In fact, the reverse can be true.
Scriptless testing takes away laborious manual testing. Everyone — test engineers, developers, business testers — can focus on other aspects of their roles. For example, test engineers can spend more time stabilizing their test automation scenarios.
Plus, scriptless testing with AI does not cover every test. Even as it evolves, likely, there will always be a need to write code for test scripts. In practice, that means finding the right balance between the two and working out which tests are suitable for scriptless or require manual intervention and more test expertise.
Finally, it will be essential to have an overall view of all types of tests, scriptless or not, so utilize the test management layer provided within testing tools to control and view all test scenarios in one place.
Scriptless AI-based testing continues to develop. Over the coming months and years, it will demand less input from testers and, ultimately, will be more autonomous. However, people are always going to be needed behind the scenes.
In the meantime, scriptless testing, if applied correctly, has some powerful benefits to offer straight away, helping the telecom industry keep up with the demand for high-quality software.
Scriptless AI-based testing for web and mobile
The challenges of testing are further amplified in mobile environments, with the need to support cross-platform development frameworks such as Flutter and React Native. Multiple OSes, devices, screen resolutions, and OS versions makes continuous testing for mobile a vital part of the software release cycle for telcos.
Fortunately, AI-based testing tools have evolved to address these challenging environments and go far beyond basic ‘play-and-record’ testing solutions. Examples of features include sensor-testing, interacting with the devices and not just with the application, fingerprint, network conditions, location, and other advanced scenarios.
These tools can also handle disruption testing (such as receiving calls during testing), fingerprint, network conditions, location, and other advanced scenarios.
Future of AI-based testing
AI-based testing continues to evolve and goes beyond just the method of test creation. Next generation tools learn the application and then model it using self-learning engines, which allow tests to be auto maintained.
Just like open-source automation, scriptless and AI-based tools use identifiers, also known as locators, to find and interact with elements. Where they differ is that these tools remove the friction of managing and spelling out these locators with a sophisticated self-selection locator system.
Sophisticated autonomous locators also expand the support to edge cases such as elements that are not in the object tree/DOM (the XML that gets rendered on the screen) and provide telcos with insights into quality through advanced reports by build, platform, and more. Some testing tools can even pre-build tests.
These are just some examples of how AI plays a role in making testing smarter, improving the productivity of teams involved and the ability to do more with less: more coverage, with fewer resources and in a shorter timeframe.
In turn, that means the telco industry can deliver better products and faster, despite the continued scaling up of devices, platforms, and versions to support.
By Shani Shoham, founder of 21 Labs, a Perforce company