MUMBAI, India, Oct. 5 -- Intellectual Property India has published a patent application (202611100441 A) filed by Manish Tiwari; and Shreya Pandey on August 19, 2026, for Consensus-Gated Unsupervised Pipeline For Rotating-Machinery Fault Detection, Characterization, And Severity Assessment From Route-Collected Vibration Data.
Inventors include Manish Tiwari; and Shreya Pandey.
The application for the patent was published on October 02, 2026, under issue no. 40/2026.
Abstract: The present invention discloses a computer-implemented, unsupervised software processing pipeline for rotating-machinery fault detection, characterisation, and severity assessment from routine route-collected vibration measurements. The system extracts 25 features driven by a single configurable parameter list of machine- specific fault-frequency multipliers and converts them into percentage deviations from an automatically computed healthy baseline. In a sequential three-stage workflow, a Stage 1 consensus detection gate evaluates measurements using an ensemble of five mathematically dissimilar novelty detectors, declaring an anomaly only when a 3-of-5 threshold is satisfied. Stage 2 executes relative-evidence fault characterisation by subtracting cross-band average growth to eliminate overall vibration bias, applying a temperature-controlled softmax to yield per-band component probabilities. Stage 3 maps individual detector outputs into a standardized 0–100 continuous health index via piecewise linear interpolation between saved baseline percentile anchors. The invention provides automated data triage, layered confidence reporting, and immediate day-one monitoring capabilities without requiring historical failure records.
Disclaimer: Curated by HT Syndication.
