Unbalance in rotating machinery is a major cause of excessive vibration, leading to reduced efficiency, increased wear, and potential mechanical failure. This paper presents a comprehensive approach for detecting unbalance in a rotating shaft using vibration signature analysis combined with Fast Fourier Transform (FFT). Vibration signals are acquired using accelerometers and analyzed in both time and frequency domains. FFT enables transformation of complex time-domain signals into frequency-domain spectra, allowing identification of dominant frequency components corresponding to shaft rotation. Experimental observations confirm the presence of a significant peak at the fundamental frequency (1× RPM), indicating unbalance. Results further demonstrate that vibration amplitude increases with rotational speed and unbalance severity. The proposed sys-tem provides an accurate, reliable, and cost-effective solution for predictive maintenance in rotating machinery across industrial applications.