CODEMIND Corporation: Fully Automated White Box Testing Through COYOTE
Seungcheol Shin, CEOThat’s where CODEMIND Corporation’s steps in.
“The tools and solutions we provide at CODEMIND are advantageous to those who are developing or/and using software and need to analyze and verify their product,” says Seungcheol Shin, the CEO of CODEMIND Corporation.
One of CODEMIND’s top products, Coyote can automatically generate test cases for C/C++, resulting in high coverage, which is incomparable to the past.
A larger number of test cases do not necessarily lead to higher test coverage; however, high-quality test cases result in high test coverage with high probability. CODEMIND applied its symbolic testing technology to Coyote, which makes white box testing fully automated and allows clients to get over 90 percent unit test coverage without having to code.
With a single click, clients can process hundreds of thousands of lines of source code in just a few hours at the longest, which is the fastest performance ever achieved.
Coyote performs the most rigorous white box testing in which every single execution logic and structure inside the software is scrutinized while all execution paths are tested. Over 90 percent of test coverage suggests that the software tested has a lower chance of containing undetected bugs, which leads to more reliable software with much higher calibre.
With Coyote, clients can experience symbolic testing technology and high-level testing technology that combines static analysis technology, methods for dynamic symbolic execution, and machine learning techniques. It can drastically reduce the enormous human and material costs associated with the testing of safety-critical software such as embedded software. Since white-box testing, even if it is with a foreign-made testing tool, requires manpower with development skills, the cost of human resources is inevitably high. It is expected to be especially beneficial for automotive electronics and electrical equipment makers that utilize foreign workers with relatively low labor costs.
Coyote was first applied to verify automotive electrics software, and as a result, productivity was shown to improve up to orders of magnitude. Its strengths are even more pronounced in the large-scale and complex autonomous vehicle software field. As the utility and effectiveness of Coyote have been proven, it is expected to be actively applied to various industries such as finance, communication, aviation, and robotics as well as the automobile field. If the scope of application and utilization of Coyote is expanded, the relatively unnoticed field of white box testing will become more active, and as a result, it is expected that the level of software verification will be further enhanced.
Coyote performs the most rigorous white box testing in which every single execution logic and structure inside the software is scrutinized while all execution paths are tested. Over 90 percent of test coverage suggests that the software tested has a lower chance of containing undetected bugs, which leads to more reliable software with much higher calibre.
With Coyote, clients can experience symbolic testing technology and high-level testing technology that combines static analysis technology, methods for dynamic symbolic execution, and machine learning techniques. It can drastically reduce the enormous human and material costs associated with the testing of safety-critical software such as embedded software. Since white-box testing, even if it is with a foreign-made testing tool, requires manpower with development skills, the cost of human resources is inevitably high. It is expected to be especially beneficial for automotive electronics and electrical equipment makers that utilize foreign workers with relatively low labor costs.
The tools and solutions we provide at CODEMIND are advantageous to those who are developing or/and using software and need to analyze and verify their product
