AI in Software Testing at Tricentis Accelerate 2018

As Artificial Intelligence (AI) continues to grow in industries such as food and beverage, music, and even beauty it’s easy to understand why software testers may be concerned about AI taking over their jobs. AI has been around for many years, but has never felt more omnipresent than now.

“It seems as though we are just a baby step away from eliminating the need for human thinking in software testing,” said Ingo Philipp, who is Distinguished Evangelist for Tricentis and recently spoke in Vienna at Tricentis Accelerate.

Philipp, who took the main stage on Oct. 9, gave one of four talks focused on AI at this year’s conference. His session, “AI in Software Testing: The Best (and Worst) Uses,” focused on exploring how AI could assist or replace the human tester in specific software testing use cases.

Generally, AI can be described as a machine with the ability to apply intelligence to any problem. In the software testing space, AI can be used to help answer the question “How can AI assist the human tester in specific testing use cases?” This is exactly the question that Philipp and his team set out to answer.

After surveying testers at 720 organizations and collecting more than 5,400 responses, Philipp and his team created 10 categories of problem areas, including:

    • Redundancy prevention, in which AI eliminates and prevents redundancies in test case portfolios to achieve the same results in terms of business risk coverage but with less effort
    • Test strategy optimization, in which AI uses production data to prioritize features, to define what to test, what to automate, and even what to build
  • Automated defect diagnosis, in which AI proposes potential reasons that caused a test case to fail to help development reduce the time it takes to analyze the root cause of a defect

Other categories included automated
 test design, risk coverage 
optimization, automated
 exploratory testing, resilient 
automation, false-positive
 detection, user experience
 analysis and portfolio 
inspection — all of which can be aided by the use of AI.

At another Tricentis Accelerate session, Sogeti’s Tom Van de ven discussed how AI might change the role of the tester. “I see the role of testing changing to [something like] a weather forecaster and looking at all these models and choosing the test set that is most likely to identify bugs,” Van de Ven said in a session titled “Predictive Testing: New Testing Strategies with AI.”

In the future, testers who can understand and apply AI algorithms and predictive modeling will become invaluable — testers who can apply AI correctly will be able to improve both the efficiency and effectiveness of their organization’s testing efforts.

Accelerate attendees also got a sneak peek at two AI testing tools. In a session titled “AI Testbot as a First Class Citizen of Continuous Testing,” Embian’s Jae-Jun Hwang demonstrated how an AI testbot could address the continuous testing pipeline from a new perspective. 

Later, Giovanni Denaro demonstrated AutoBlackTest (ABT), a tool that generates system test cases for interactive applications.

While testing might change the way testers approach their work, it should be seen as an enhancement, and testers shouldn’t fear being replaced by a bot. As Philipp wrapped up his presentation he left us with a quote from Jerry Weinberg: “The number one testing tool is not the computer, it is still the human brain — so don’t expect AI to solve all your problems soon.”