MUMBAI, India, Oct. 26 -- Intellectual Property India has published a patent application (202611052129 A) filed by National Institute Of Technology, Delhi; and Coresysmeta Private Limited on April 24, 2026, for Federated Learning Enabled Privacy-Preserving System And Method For Digital Twin-Based Traffic Congestion Prediction And Adaptive Traffic Management.
Inventors include Chetan Sarup Mishra; Anurag Singh; and Ujjwal Kumar Kalla.
The application for the patent was published on October 23, 2026, under issue no. 43/2026.
Abstract: Federated Learning-Enabled Privacy-Preserving System and Method for Digital Twin-Based Traffic Congestion Prediction and Adaptive Traffic Management The present disclosure relates to a federated learning-enabled privacy-preserving multi-sensor edge-based system (100) and method (200) for traffic congestion prediction and adaptive traffic management at geographically distributed traffic intersections. The system (100) comprises edge computing nodes (102), multi-sensor acquisition modules (104), probabilistic sensor fusion engine (114), a transformer processing module (116), a digital twin engine (120), a model degradation module (122), and a signal optimization module (124). The multi-sensor acquisition module (104) acquires traffic data through a camera (106), radar sensor (108), acoustic sensor (110), and vibration sensor (112). The digital twin engine (120) determines predicted traffic conditions for adaptive traffic signal control. Federated client modules (126) generate updates from locally available traffic data, which are encrypted and communicated to a central aggregation server (130) for secure aggregation and distribution. Raw traffic data remains at the respective edge computing nodes (102), providing privacy-preserving distributed traffic prediction and adaptive traffic management.
Disclaimer: Curated by HT Syndication.
