MUMBAI, India, Aug. 12 -- Intellectual Property India has published a patent application (202621064162 A) filed by Prof. Roshani Pawar; Tejas More; Atharv Mulik; Siddhi Naik; and Vaishnavi Nile on May 21, 2026, for Ai-Powered System For Smart Urban Issue Redressal..
Inventors include Prof. Roshani Pawar; Tejas More; Atharv Mulik; Siddhi Naik; and Vaishnavi Nile.
The application for the patent was published on August 07, 2026, under issue no. 32/2026.
Abstract: The present invention relates to an AI-powered multi-modal urban grievance redressal system for intelligent classification, prioritization, and routing of civic complaints in smart city environments. The invention integrates Natural Language Processing and Deep Learning techniques to process both textual and image-based complaints submitted by citizens through a digital platform. The textual complaints are classified using TF-IDF feature extraction and Logistic Regression algorithms, while image-based complaints are analyzed using a MobileNetV2 deep learning model trained through transfer learning techniques. The invention further incorporates a decision-level weighted fusion mechanism that combines outputs from text and image classification models to improve classification robustness and accuracy. The system additionally includes geo-tagging, automated complaint prioritization, real-time routing to relevant municipal departments, and analytical dashboards for monitoring complaint trends and departmental performance. The invention is implemented using a scalable architecture comprising Flutter-based mobile applications, Fast API backend services, PostgreSQL database management, and Docker containerization for cloud deployment. Experimental evaluation demonstrates improved complaint classification performance and reduced grievance processing time compared to traditional manual systems. The invention enhances transparency, accountability, citizen engagement, and operational efficiency in urban governance and smart city administration system
Disclaimer: Curated by HT Syndication.
