Sunil Tyagi, PhD

Working note · Signal processing

DEMON Analysis: Reading a Vessel’s Propulsion from Its Own Noise

A classical technique that still earns its place alongside deep learning — and one way it caught a bug that spectrograms missed.

Read the full note (PDF)
DEMON analysis output: shaft- and blade-rate harmonics extracted from the cavitation envelope of a vessel recording Blade-rate harmonics extracted from the cavitation envelope of a vessel recording — output from Oravont’s own DEMON tool.

What the note covers

A cavitating propeller is a broadband noise source, but the cavitation is not steady: each blade loads and unloads as it sweeps through the wake, amplitude-modulating the broadband noise at the blade-passage rhythm. The broadband cavitation is the carrier; the propeller kinematics are the modulation. DEMON — Detection of Envelope Modulation on Noise — demodulates the carrier and reads the rhythm, recovering shaft rate and blade rate from noise alone.

The note is a short, practical walk-through: what DEMON extracts, the four-step pipeline (bandpass → Hilbert envelope → AC-coupling → envelope FFT), how to read the output, the band-selection decision most tutorials skip, and a case study where DEMON exposed a defect in a synthetic dataset that conventional spectrograms had hidden. Everything is textbook-level physics applied to open material — deliberately, because the value of DEMON is precisely that it is transparent.

Full note (PDF) · Also published on LinkedIn · Oravont Systems · 2026