Spectrum Analysis Service - Developer Documentation - Developer Documentation
Skip to content

Spectrum Analysis Service

Idea

Spectrum Signal Analysis allows users to perform time domain and frequency domain analysis. It provides functions to transform a time-domain signal into its frequency components (via Discrete Fourier Transform) and to detect threshold breaches of their amplitudes.

Access

For accessing this service you need to have the respective roles listed in Spectrum Analysis roles and scopes.

Basics

Based on sound emitted by industrial equipment and components one can assess equipment’s health status and/or expected life duration.

Discrete Fourier Transform

The Discrete Fourier Transform (DFT) employs the Fast Fourier Transform to transform given audio data from time domain to frequency domain. Each frequency is represented by a sinusoidal oscillation that has its own amplitude and phase. This transformation allows the user to analyze the signal for certain patterns or for threshold detection. Before transforming the data into frequency domain, it is processed using a window function to reduce spectral leakage.

Features

The Spectrum Analysis Service exposes its API for realizing the following tasks:

  • Transform signals from time domain into frequency domain via Discrete Fourier Transform
  • Detect threshold violations

Limitations

The input signal must be of .wav format with a sample size of 8 or 16 bits and a maximum file size of 1 MB.

Example Scenario

The service engineer in charge of maintenance for multiple automated production lines wants to have an early signal when a certain mechanical component is about to fail. They install a sound recorder to send 1 MB slices to Insights Hub for analysis.

They use the Spectrum Analysis Service to transform the signal into frequency domain and forward the results to the Visual Flow Creator to detect frequency intervals of interest. The result could for example be, that a certain bearing reaches a resonance condition and produces a specific sound, shortly before a machine is about to fail.

Except where otherwise noted, content on this site is licensed under the Development License Agreement.

Last update: March 17, 2023