Abstract— This study examines the degradation of known positioning data in Ultra-Wideband (UWB) Real-Time Location Systems (RTLS) under Non-Line-of-Sight (NLOS) conditions, which are frequently encountered in indoor environments (e.g., walls, metal cabinets, shelves). The Decawave DW1000, which uses Two-Way Ranging (TWR), demonstrates reliable performance in Line-of-Sight (LOS) conditions, but in NLOS conditions, multipath causes significant deviation and dispersion. To mitigate these effects, we propose an IMU-assisted fusion approach that integrates a 6-axis Inertial Measurement Unit (IMU) with the UWB pipeline. This approach reduces the NLOS positioning error by up to 89.52% compared to the baseline using UWB alone. With the proposed method, the Mahony filter is used with PID gains (K𝑝, K𝑖) tuned for real-time response; accelerometer cues detect and eliminate sudden jumps caused by multipath. The combined UWB-IMU measurements are then processed by an Extended Kalman Filter (EKF), which explicitly models the temporal dynamics and measurement uncertainty, yielding smoother and more reliable data. We evaluate the method under LOS and NLOS conditions against a baseline system using only UWB; MATLAB-based analyses confirm the stated improvement. These findings validate the effectiveness of IMU-assisted fusion for UWB-based RTLS in complex indoor environments and present a practical method for high-accuracy positioning with modest computational load and sensor complexity.