MUMBAI, India, Aug. 31 -- Intellectual Property India has published a patent application (202621084329 A) filed by National Institute Of Technology Raipur on July 09, 2026, for A Two-Phase Embedded System For Road Anomaly Detection Using Adaptive Signal Processing And Machine Learnin.
Inventors include Dr. Suman Kumar Saha; and Dr. Sukeshni Tirkey.
The application for the patent was published on August 28, 2026, under issue no. 35/2026.
Abstract: The present invention relates to a two-phase embedded system (100) for road anomaly detection using adaptive signal processing and machine learning. The system (100) comprises a primary three-axis accelerometer (1) for acquiring road-induced vibration data, primary preprocessing modules (2, 3, 4), a secondary three-axis accelerometer (5) for acquiring reference vibration data, and secondary preprocessing modules (6, 7, 8). The system further comprises adaptive filter modules (9, 10, 11) configured to perform per-axis adaptive noise cancellation using a Metaheuristic/LMS algorithm. A filtered road condition data generation module (12) generates filtered road condition data from denoised multi-axis signals. A manual annotation module (13) provides labelled road-event data to a machine learning pipeline (14), which trains or executes a road condition classification model. A classified road condition output module (15) generates road condition outputs including smooth road, bump, pothole and rough road conditions. The invention enables real-time, low-latency and noise-resilient road anomaly detection on a vehicle-mounted embedded platform.
Disclaimer: Curated by HT Syndication.
