<![CDATA[This study presents the design and implementation of a Kalman-Bucy filter for fault detection in DC motor systems, which are widely used in industrial drives and automation. Accurate state estimation is essential for ensuring reliable operation, particularly in the presence of measurement noise and parameter uncertainties. The proposed observer exhibits rapid convergence in speed estimation (less than one second) and strong robustness to measurement noise, achieving a Root Mean Square Error (RMSE) of 24.38 rad/s, closely matching the noise standard deviation (σᵥ = 23.01 rad/s). This close agreement indicates that the Kalman-Bucy filter operates near its theoretical optimal performance under Gaussian noise assumptions. Fault detection is carried out through residual analysis under three fault scenarios: ramp, inverse ramp, and square wave. Each scenario generates distinct residual patterns, providing clear indicators of both gradual and abrupt anomalies. Quantitative evaluation demonstrates high sensitivity (97.0% for ramp and inverse ramp, 94.1% for square), perfect specificity (100%), and a zero false alarm rate across all scenarios. These findings highlight the potential of the Kalman-Bucy filter as a reliable and computationally efficient approach for state estimation and fault indication using data representative of a real DC motor system. The results provide a valuable basis for developing predictive maintenance strategies and improving system reliability. Future work will focus on experimental implementation and validation to confirm its performance under real-world operating conditions.]]>
