The partnership between one major player in the field of railway signaling and integrated transport system and PSA was established through a successful service relationship. Since 2011, PSA has provided scalable design, IT, and service support to the client through projects dedicated to the signalling and automation of conventional and high speed metro networks and railway/tram lines for transporting both passengers and goods. This also included VectorCAST testing of the Automatic Train Protection (ATP) subsystem.
All of the projects related to VectorCAST involved two main categories: C++ unit testing for coverage purposes and C/C++ requirements-based testing.
In accordance with CENELEC 50128, PSA conducted Unit regression testing based on code design and implementation. PSA was also required to perform tests in order to identify the differences between the original and updated source code. This was accomplished by creating test unit cases.
Within an aggressive timeline, PSA was able to align the quality requirements of the client with strict industry standards. This demonstrated that PSA can establish an integrated and shared partnership model, while incorporating a high level of international standards for railway safety software.
The VectorCast embedded software testing platform was used in this project to help automate both unit and integration testing. PSA chose this TÜV-certified tool for C/C++ because it ensured safe, reliable, trustworthy results, showed statistics, and supported various compilers. The tests created and performed by our QA team complied with SIL 3-4, which was verified by further independent audits.
The scope of work was often divided into several phases to verify that the identified issues had been resolved during previous testing and to ensure the delivery of a perfect quality product. Aside from actual testing, engineers performed test cases definition based on the requirements documentation.
PSA provided testing for various projects for this customer, such as the Ui-Sinseol ATP Project, NYCT, Miami ATP, LIRR 250, ViPro 1.00Q, LIRR 1500, covering a wide geographic area in North America, Europe and China.
At the beginning of the project, our team faced a number of challenges that promoted a steep learning curve. The issues included manual memory allocation and memory map creation for the simulated operation environment and building of the VectorCAST environment using various compilers, etc. As a result of these challenges, our team acquired new knowledge and developed new competences.
- Classic unit testing.
- Integration testing.
- ATP code regression testing.
- Test cases description.
- Problem report documentation.
- 7 Team Members
- 6 Months