MUMBAI, India, Jan. 2 -- Intellectual Property India has published a patent application (202541122462 A) filed by Paavai Engineering College, Namakkal, Tamil Nadu, on Dec. 5, 2025, for 'context-aware smart home automation using reinforcement learning and iot sensors.'
Inventor(s) include Dr. R. Mohanapriya; Dr. M. Shantha Kumar; Dr. A. Parimala Gandhi; Dr. M. Udhayavani; Mrs. S. Vijaya; Mrs. L. Sujitha; Mrs. A. Sujitha; Mrs. A. Aarthi; and Mr. S. Vijay Murugan.
The application for the patent was published on Jan. 2, under issue no. 01/2026.
According to the abstract released by the Intellectual Property India: "The present invention provides a context-aware smart home automation system that integrates IoT sensors with machine learning to autonomously optimize energy usage, comfort, and safety. The system employs a Distributed IoT Environmental Sensing Technique (DIEST) to collect multi-modal environmental and occupancy data from a plurality of smart home sensors. A Multi-Modal Sensor Fusion Technique (MSFT) and a Dynamic Context Inference Method (DCIM) are used to process the sensor data to generate a high-level Context State Vector (CSV) representing user activity, environmental conditions, and temporal factors. A Context-Aware Adaptive Reinforcement Learning (CAARL) Method, implementing the Home Environment Optimization Algorithm (HEOA), determines optimal control actions for home devices. A reward model, defined by an Energy-Comfort Optimization Reward Model (ECORM), enables continuous self-learning based on energy efficiency, comfort metrics, and user override events. Device commands are executed through an Adaptive Device Control Technique (ADCT) utilizing an Action Translation and Dispatching Method (ATDM). User feedback is incorporated via a User Preference Feedback Adaptation Method (UPFAM) to refine learning and personalization."
Disclaimer: Curated by HT Syndication.
