Fast Fourier Transform (FFT)

The fast Fourier transform is an efficient algorithm for computing the discrete Fourier transform, which converts a sampled time signal into its frequency components; in condition monitoring it produces the spectra used to diagnose machine faults.

A vibration or current signal measured over time is a mixture of components at different frequencies. The discrete Fourier transform decomposes a block of samples into amplitudes and phases at evenly spaced frequencies, and the FFT computes the same result far faster by exploiting symmetries in the calculation, most famously in the algorithm published by James Cooley and John Tukey in 1965. The resulting spectrum shows peaks at frequencies related to running speed, gear meshing, blade passing, the electrical supply and bearing defects.

FFT analysers and software are standard in vibration analysis, motor current signature analysis, acoustic measurement and many other fields. Analysts relate spectral peaks to machine components: a peak at running speed often indicates unbalance, gear mesh frequency equals the number of teeth multiplied by shaft speed, and bearing defect frequencies depend on bearing geometry. Spectra are trended over time to detect developing faults.

Several settings determine the quality of a spectrum. The sampling rate must exceed twice the highest frequency of interest, according to the Nyquist criterion, and anti-aliasing filters prevent false peaks. Frequency resolution equals the sampling rate divided by the number of samples, so finer resolution needs longer measurements. Window functions such as the Hanning window reduce leakage, averaging reduces random noise, and machines with varying speed may need order tracking.

Key points

Efficiently computes the discrete Fourier transform of sampled signals.
Popularised by the Cooley-Tukey algorithm published in 1965.
Turns vibration and current signals into spectra for fault diagnosis.
Sampling rate, resolution, windowing and averaging affect spectrum quality.

Where AiVibe comes in

AiAmbA AI Factory use cases include predictive maintenance and visual inspection on textile, automotive and electronics lines. AiAmbA IoT normalises industrial signals from protocols such as OPC UA and MQTT, with driver availability confirmed per installation.

Explore AiVibe’s work in Quality, Reliability & Maintenance →

Fast Fourier Transform (FFT)

Key points

Where AiVibe comes in

Related terms