Worked on the EPFLXplore/ERC_HD repository to enhance the reliability of trajectory planning in robotic systems using C++. Focused on addressing a critical edge case where identical initial and target joint states could lead to division-by-zero errors and NaN commands. Implemented a guard within the trajectory computation logic to skip calculations when no movement is required, thereby improving system robustness for both simulation and real-world applications. This fix reduced the risk of undefined or unstable commands, contributing to safer and more maintainable control software. The work demonstrated attention to detail in robotics, trajectory planning, and C++ development practices.