MUMBAI, India, July 11 -- Intellectual Property India has published a patent application (202541060990 A) filed by Anusha Nerella; and Pratik Badri, Wyra, Telangana, on June 26, for 'real-time fraud detection in digital payments using federated learning and iot sensors.'
Inventor(s) include Anusha Nerella; and Pratik Badri.
The application for the patent was published on July 11, under issue no. 28/2025.
According to the abstract released by the Intellectual Property India: "The present invention proposes a novel system for real-time fraud detection in digital payments by integrating federated learning algorithms with IoT (Internet of Things) sensor networks embedded in edge payment devices. The invention addresses the growing concern over digital payment fraud, especially in an era where financial transactions are increasingly contactless, mobile-driven, and decentralized. Traditional fraud detection mechanisms often rely on centralized data collection and analysis, posing risks related to data breaches, delayed responses, and limited adaptability to evolving fraud patterns. This invention overcomes these challenges by offering a privacypreserving, distributed, and intelligent fraud detection framework. At the core of the system is a federated learning engine deployed across a network of IoT-enabled devices such as smartphones, POS terminals, smart cards, and wearable payment systems. These devices locally train lightweight machine learning models using real-time user interaction data-such as transaction speed, location, typing behavior, device motion, and environmental cues. Rather than transferring raw personal data to a central server, the system transmits only encrypted model updates, thereby ensuring data security and regulatory compliance with standards like GDPR and India's DPDPA. An intelligent central aggregator combines these encrypted model updates to build a refined global fraud detection model, which is then redistributed to all participating devices. This continual learning process ensures that the system evolves dynamically in response to emerging fraud tactics. Simultaneously, IoT sensors on the devices perform real-time contextual analysis of transactions, detecting anomalies based on behavior and environmental mismatch, and enabling instant action-such as flagging, blocking, or requiring verification. This invention provides a scalable, responsive, and secure solution for stakeholders including banks, mobile wallet providers, fintech platforms, and e-commerce vendors. By combining federated learning with sensor-driven edge intelligence, it delivers robust protection against fraud, maintains user privacy, and strengthens trust in the global digital payment ecosystem."
Disclaimer: Curated by HT Syndication.
