This thesis explores the essential role of unit testing in refactoring the Qualcomm Cloud AI 100 firmware based on an RTOS environment. The project presented several challenges, particularly in ensuring the system's reliability and robustness throughout the refactoring process. To address these issues, a comprehensive unit test suite was developed, focusing on validating the core functionalities of the system while maintaining its structural integrity. By adhering to the four pillars of effective unit testing—protection against regression, resistance to refactoring, fast feedback, and maintainability—the approach enabled thorough testing of key components. The resulting testing framework provides a strong foundation for long-term maintenance and system scalability.