Industrial robots are the crucial system for manufacturing automation and smart factory. However, the unexpected failures of the industrial robots cause a large amount of downtime and loss in the production line. Thus, condition monitoring system for the industrial robots receives considerable attention in recent years. This paper presents fault detection methods of an industrial robot. For robust fault detection under various operating conditions, two methods are developed based on acceleration signal and control signal. The performance of the proposed methods is demonstrated using a 6-DOF industrial robot with normal and faulty cases under various operating conditions. The method based on acceleration signal shows the robust performance in terms of rotating speed, loading condition and different postures of robot arms.