MUMBAI, India, July 7 -- Intellectual Property India has published a patent application (202641080574 A) filed by Narsimha Reddy Engineering College Nrcm on June 30, 2026, for Reinforcement Learning-Based Autonomous Traffic Signal Optimization System For Smart Cities.
Inventors include Srilakshmi Cherukuri; Pallela Dileep Kumar Reddy; Gaddam Anantham; Bandi Raju; R. Jeevitha; and B. Chaitanya.
The application for the patent was published on July 03, 2026, under issue no. 27/2026.
Abstract: A deterministic multi-agent reinforcement learning-based autonomous traffic signal optimization system for smart city intersections is disclosed. The system comprises an intersection sensing module (210), a signal conditioning and data acquisition module (220), a deterministic embedded signal controller unit (230) including volatile memory (231) and non-volatile memory (232), a reinforcement learning inference engine (240), an inter-agent communication module (250), a traffic signal actuation module (260), and a supervisory safety enforcement module (270). Real-time vehicular state parameters including queue length, vehicle speed, and occupancy are acquired and digitized to construct structured state vectors. A reinforcement learning policy computes phase selection actions, while adjacent intersections exchange compressed state embeddings through vehicle-to-infrastructure communication. A hardware-level supervisory module validates each phase transition against statutory safety constraints. All sensing, inference, cooperative exchange, and actuation operations are executed within a deterministic processing cycle below 100 milliseconds, thereby enabling distributed cooperative signal optimization with bounded latency and improved congestion reduction.
Disclaimer: Curated by HT Syndication.
