MUMBAI, India, Aug. 31 -- Intellectual Property India has published a patent application (202641099437 A) filed by Rajalakshmi Institute Of Technology on August 18, 2026, for Machine Learning-Based Autonomous Traffic Flow Optimization System.
Inventors include Mr. Rajesh C; Ms. Geetha A; Ms. Avudai Selvi S; and Ms. Gajalakshmi R.
The application for the patent was published on August 28, 2026, under issue no. 35/2026.
Abstract: The present invention relates to an autonomous traffic flow optimization system configured to continuously observe, predict, evaluate, and regulate traffic conditions across one or more interconnected transportation corridors using machine learning and adaptive control techniques. The system integrates heterogeneous traffic information obtained from roadside sensing units, imaging devices, connected vehicles, traffic signal controllers, road infrastructure sensors, navigation-related information, environmental sensors, and historical traffic records to construct a continuously updated representation of traffic conditions. A machine learning processing architecture analyzes the acquired information to identify traffic density, vehicle arrival patterns, queue formation, lane utilization, congestion propagation, abnormal traffic conditions, and anticipated traffic demand. Based on the generated traffic state representation and predicted future conditions, an optimization engine determines traffic control actions including signal phase selection, green-time allocation, phase sequencing, cycle duration, offset adjustment, lane utilization recommendations, diversion strategies, and coordinated corridor control. The system evaluates the anticipated effect of each control action before implementation and autonomously selects a control action according to one or more optimization objectives including reduction of vehicle delay, queue length, travel time, stop frequency, congestion propagation, fuel consumption, and emissions. A feedback mechanism continuously compares predicted traffic behavior with subsequently observed traffic behavior and updates one or more machine learning models and optimization parameters. The system thereby provides adaptive, predictive, and autonomous traffic management capable of responding to changing traffic demand, incidents, special events, environmental conditions, and other dynamic transportation conditions without relying exclusively on predetermined signal timing plans.
Disclaimer: Curated by HT Syndication.
