Worked on enhancing heading accuracy and reliability in the PhotonVision/photonvision repository by developing and shipping a dedicated Heading Data Reset API for the PhotonPoseEstimator. This feature enables the system to clear existing heading data and reseed the buffer with fresh data following pose or rotation resets, which improves the stability of heading calculations in dynamic robotics environments. The approach focused on reducing drift and supporting more accurate downstream tracking and navigation. Leveraged expertise in computer vision, embedded systems, and robotics, implementing the solution in Java to deliver cleaner heading computations and contribute to more robust live deployment performance.