“Deep Compatibility Testing” service launched today!

Smartphones account for about 55 percent of all users, or about 2.75 billion people who own one or more smartphones, according to Ericsson’s most recent publicly available report. Android software compatibility has also become an important standard for users to measure product quality in the Internet + era.

However, at the present stage, the new Android models in the market are fast, and the average number of mobile products released annually is still as high as 2,000. At the same time, the system upgrade of these models is fast and the fragmentation is serious. Based on hundreds of millions of mobile users, the difference of users covered by different models may be calculated in tens of thousands for application positioning.

In addition, the cost of deploying TOP300 phones is as high as 3 + MILLION yuan, and the cost of deploying a test team of 4-5 people is also over 1 million yuan. It is difficult for companies to get all the mainstream models on the market for testing. Coupled with the lack of professional talents, precision standard and efficient management tools, companies cannot spend more budgets, make developers cannot to mainstream terminal market models to find the adapter device testing, cause of software compatibility, poor display, congruent a series of problems, plus can’t first response to user feedback, word of mouth decline eventually face failure.

In order to reduce the testing cost of enterprises, Tencent WeTest is determined to introduce AI and automation capabilities.

AI, a scientific and efficient artificial intelligence, is penetrating into every aspect of people’s lives. In the front line of software development, AI also has its place in order to solve the developers’ four major difficulties of “lack of equipment”, “lack of manpower”, “lack of budget” and “lack of mature testing methods” and ensure the efficient production and safety of product content.

Recently, WeTest, Tencent’s one-stop quality open quality platform, launched a new service of deep compatibility testing based on AI automation, and upgraded all its testing functions to support developers to solve testing problems within the full software life cycle.

Tencent WeTest integrates AI functions in the two modules of simulation automation and detection and analysis to provide tailored services for different applications, freeing manpower to the greatest extent and improving test efficiency.

AI automation testing overall framework

• Control and API interface automation, effectively obtain state information to test twice the result with half the effort

  
  

For mobile games, WeTest launched the automation framework GAutomator. Through this framework, testers can identify the control elements on the game interface directly. Relying on the RPC mechanism of GAutomator, testers can call the private API interface provided in the game at any time and get the game state information in the first time. Specific processing is accomplished through API.

And in view of the application of APP, breakthrough the traditional way of scripting, based on the identification and control the double object recognition mechanism of intelligent image recognition, through automatic exploration test framework based on the AI and recording tool, the tester only need to work again on a mobile phone can be directly obtained key scene model, and then complete the application under test each cover and exploration.

• Image recognition realizes AI automation and supports deep scene exploration

Of course, AI automation exploration is smart and efficient, but in testing, how high can be achieved in more subcategories of application is one of the questions developers need to measure.

In terms of AI automation exploration, image bug identification has become the biggest advantage of AI technology in the test link. For example, in application compatibility testing, using product samples, AI can detect black screen, white screen, splintered screen and other problems.

Automatic AI processing process for image bug problems

Based on the field of games, Tencent WeTest tries to conduct effective scene exploration and selection operation through the combination of OCR technology and AI. Through the classification of game UI semantics and ICONS in a more complex game scene to understand and identify, combining the recognition results, to obtain the name and location of the current UI image corresponding to the operational elements, and then explore the scene more truly.

UI image recognition and understanding and automatic running examples

Part of the template matches the database sample image

Tencent WeTest AI automation capability has been verified in a number of internal and external products, and the compatibility efficiency has been comprehensively improved. In the deep compatibility testing service, we covered the core scenarios of products for users based on AI automation ability, and conducted compatibility tests on the mainstream TOP300 terminals in the market.

Deep compatibility with service testing service processes

Deep compatibility service problem reporting case

In addition to basic services, the integration of AI automation and deeply compatible service testing functions, as well as the realization of cloud games, AI games, AI automated testing, AI-based synchronous control system, performance big data analysis, public opinion big data analysis, public opinion early warning and other cutting-edge technical support, Automated testing, crowdsourcing testing, user experience testing, APM monitoring, public opinion analysis and other services will be implemented in real life to achieve a 1+1>2 effect in terms of testing efficiency and quality.

Ten years ago, Tencent began to accumulate experience in testing methodology, quality system and testing tools based on its own game products.

With the promotion of Tencent’s opening-up and ecological strategy, Tencent has introduced many games and applications from industry partners. In order to ensure that all products launched on Tencent platform have the basic quality in line with the industry and Tencent standards, WeTest platform was officially opened to all developers in 2015. Also gradually from the accumulated experience from the internal verification to the industry verification of the service model, so as to maximize the quality of the developers continue to move forward.

At present, Tencent WeTest deep compatibility test has been officially launched, providing monthly package and seasonal package service discounts for individual users and enterprise users. The average discount package is as low as 5.5 yuan per model, which further solves the dilemma faced by developers with low price and high quality and promotes the development of the whole industry.

In the future, WeTest will rely on AI testing, cloud platform, big data analysis and other cutting-edge technologies to create a quality control service ecosystem, empower developers and help enterprises create high-quality products.

WeTest will continue to explore more application scenarios of AI in the field of product quality, driving industry progress with cutting-edge technology; Focusing on the whole product life cycle of “demand – development – release – operation and maintenance – feedback”, in-depth construction of code quality, layered automation, full link performance, multi-level security protection capabilities, to provide a one-stop quality assurance platform in line with the needs of the industry and enterprises for the vast number of users, continuous improvement of product quality; By monitoring and analyzing the online multi-dimensional user experience and quality big data, it provides the quality baseline and standard for the industry and promotes the benign development of the industry.

In the past few years, WeTest has cooperated with many world-renowned operating system, mobile phone, software and hardware manufacturers to provide innovative technology applications for enterprises. In the future, we hope to cooperate with more partners and enterprises committed to creating high-quality products and make progress together. Drive growth by quality and create a new quality future.