Intellectual Property India Publishes Patent Application for 'A System And Method For Adaptive Multi-Agent Traffic Signal Control Utilizing Graph Neural Networks, Real-Time Computer Vision, And Reinforcement Learning' Filed by Graphic Era Hill University, Dehradun; and Graphic Era

MUMBAI, India, Feb. 6 -- Intellectual Property India has published a patent application (202511132246 A) filed by Graphic Era Hill University, Dehradun; and Graphic Era, Dehradun, Uttarakhand, on Dec. 26, 2025, for 'a system and method for adaptive multi-agent traffic signal control utilizing graph neural networks, real-time computer vision, and reinforcement learning.'

Inventor(s) include Rajat Pundir; Ankit Negi; Sujal Thapa; Dr. Chandradeep Bhatt; and Dr. Anupam Singh.

The application for the patent was published on Feb. 6, under issue no. 06/2026.

According to the abstract released by the Intellectual Property India: "The invention relates to intelligent traffic management and comprises a perception module with IoT-enabled cameras and edge processing units executing deep convolutional neural networks for object detection and tracking, thereby extracting traffic state variables. A graph modelling module converts a road network into a directed graph utilizing a graph constructor, embedding layers, and a Graph Attention Network (GAT) with multi-head attention, dynamic coefficient computation using learnable transformations, LeakyReLU activation, and softmax normalization. A decision-making module, featuring deep reinforcement learning agents based on Multi-Agent Proximal Policy Optimization (MAPPO) and incorporating an action masking mechanism, selects optimal traffic signal phases at each intersection. The system further integrates simulation, validation, and deployment infrastructures for comprehensive real-time traffic control."

Disclaimer: Curated by HT Syndication.