Abstract: A dynamic fault model for DC-motor actuators of robot is set up. The faults of the DC torque constant, the armature resistance, the armature inductance, etc., can be described by this fault model. A dynamic fault model for sensors of, industrial robot is also sct up. Based on these models the systematic algorithm for the real-time detection and diagnosis of parameter bias faults for a general class of nonlinear time-varying stochastic systems^12,13,is successfully used to detect and diagnose these faults. Finally, the effectivess of the proposed method is demonstrated by computer simulation.