As with almost every other sector in the IT space, generative AI is rapidly making its way into the realm of mobile application development.
A survey this month by Kobiton of more than 100 developers and testers found that the majority are using generative AI tools for a range of tasks, helping them address challenges in developer and QA operations, particularly limited financial resources, inefficient development and a lack of skills in the labor market.
All of these lead to slow mobile app release cycles, which is costing organizations a lot of money. According to Kobiton’s survey, 75% of respondents said slow release cycles are costing companies at least $100,000 a year, and 13% put that number between $1 million and $10 million.
None of this is good news for companies building mobile apps, which already is more difficult than developing web apps for a number of reasons, according to Kobiton CTO Frank Moyer.
Mobile app developers have to manage a complex environment that includes myriad devices, operating systems and networks that have to be factored into design, coding and testing, Moyer told Techstrong.ai.
“Additionally, mobile script automation is much more difficult due to the unique way Android and iOS operating systems function,” he said. “Developers must optimize code, graphics and other assets to ensure smooth performance on mobile devices. Mobile apps often need to account for variable network conditions, including slow or unreliable connections. Test permutations are endless.”
There also are key differences in the development and approval processes, because unlike web developers, those working on mobile apps don’t have direct control over the deployment of their creations. Instead, they need to navigate the specific requirements and approval processes of app stores like Google Play and Apple’s App Store.
“This adds a layer of complexity and consideration in the development cycle,” Moyer said.
Throw in other issues, like the need to integrate with platform-specific services like Apple Pay and Google Wallet and variable network conditions, and the development of mobile apps gets wildly complex.
Using Generative AI for Mobile Development
Developers are adopting generative AI address the hurdles. According to Kobiton’s survey, 60% of respondents are using generative AI tools in their QA cycles to update scripts or code and 55% are using them to analyze test results. Another 47% are generating test scripts with generative AI.
In addition, 50% said they believe AI-driven automation can replace manual testing for mobile apps and 34% are excited about AI’s potential to increase the productivity of software development.
In particular, developers said AI capabilities would help their organizations’ strategy for automating testing mobile testing in such areas as predictive analytics for forecasting potential defects (51%), generating test cases and data (45%), and to better test case documentation through the uses of natural language processing (44%).
Excitement is Spreading
Kobiton, founded in 2016, offers a mobile device testing platform that the company says drives a 10-fold increase in mobile app delivery over competitors, listing AI-augmented testing as one of the features. The platform is used by high-growth startups and enterprises in such industries as retail, gaming, financial services, and travel.
The company isn’t alone in advocating for mobile app development teams and testers using generative AI tools. Custom software development company Peerbits, which also creates mobile apps, wrote in July that “with no doubt, AI technologies have ushered in a transformative era in mobile app development, revolutionizing the entire development process. With AI integration in apps, developers can unlock new levels of efficiency, productivity and innovation.”
Another custom software maker, mDevelopers, wrote in October that, “from voice recognition systems to machine learning algorithms, mobile app developers have access to novel tools enabling them craft effective solutions with improved [user interface and user experience] design outcomes than ever before, thanks largely due partly from now accessible advanced artificial intelligence powered capabilities.”
That said, there were some worries expressed in Kobiton’s survey, such as 22% of respondents believing AI might hurt software quality while 20% said AI tools could impact career opportunities for developers and testers.
Moyer said that likely indicates that generative AI is in the early stages, though he said it’s less of an emerging trends and more a sea change, including in app development. The technology can free up developers and testers from time-consuming, repetitive, and tedious tasks like manual training, which will free them up for more interesting and challenging work.
He said Kobiton’s survey results indicate that most mobile app developers are optimistic about AI.
“They’re actively exploring ways to use AI in their work, recognizing the need to speed up app delivery despite resource constraints,” Moyer said. “AI is seen as a game-changer, giving human testers valuable ‘superpowers.’ We expect this trend to gain momentum in the next few years, with the mobile app development community embracing and integrating AI solutions to stay innovative.”