In this project, I implemented Extended Kalman Filter (EKF) and Unscented Kalman Filter (UKF) algorithms for landmark-based robot localization. The goal was to estimate the robot’s position and orientation using noisy bearing measurements from known landmarks and odometry data. I developed the EKF and UKF update functions, integrated them into a simulation environment, and analyzed their performance under varying noise conditions. Additionally, I generated visualizations of the robot’s estimated trajectory, covariance ellipsoids, and pose error over time, ensuring the filters’ accuracy and robustness. This project provided hands-on experience with state estimation techniques and their practical applications in robotics.