The objective of the proposed research is to present a methodology for incipient bearing fault detection using ultrasound technology. The final objective of this research is to develop a methodology based on sensor fusion technology to detect bearing faults at an earlier stage in squirrel cage induction machines. This method is expected to be more reliable compared to conventional methods used in monitoring the bearing health. In this research, bearing faults are experimentally generated using electrolytic corrosion method and an ultrasonic microphone is used to capture the sound signals generated from the motor. In the first stage of research, the relevant features of the ultrasound data are extracted and compared with vibration data which is considered as the reference. The experiments are performed at different load conditions and the relevant fault signatures are extracted. In the second stage, fault signatures from current and stray flux sensors will be extracted and compared with the results of ultrasound and vibration sensors. In the final stage of research, a novel fault index will be developed using ultrasound, vibration, current and stray flux sensors.