Bearings are critical components in various industrial machinery, and their failure can lead to significant downtime and maintenance costs. Effective fault diagnosis methods are essential to ensure reliable operation and prevent unexpected failures. Traditional approaches rely heavily on vibration analysis, however, recent advancements in sensor technology have highlighted the potential of combining vibration data with sound data for enhanced fault diagnosis. This paper presents an investigation into diagnosing bearing faults by utilizing both sound and vibration data for employing Machine Learning (ML) algorithms. The proposed methodology integrates data acquisition, feature extraction, and ML based classification for bearing fault diagnosis. Results indicate that the fusion of sound and vibration data significantly improves diagnostic performance compared to using either data type alone, offering a robust solution for predictive maintenance in industrial applications. This research underscores the potential of multi-sensory data integration in advancing fault diagnosis technologies, paving the way for more reliable and efficient maintenance strategies.