MUMBAI, India, Feb. 6 -- Intellectual Property India has published a patent application (202541085131 A) filed by Mr. N. Darwin; Dr. Shyam Kishore. G; Dr. Jongoni Srikanth; Ms. Shobana D; Dr M. Jeyaselvi; Dr. Vijay Vasanth A; Ms. D. Linett Sophia; Dr. B. Mahesh; Dr. M. Shanmuganathan; and Dr. Jeba Johannah J, Chennai, Tamil Nadu, on Sept. 8, 2025, for 'ai-enabled low power architecture for next generation iot devices.'
Inventor(s) include Mr. N. Darwin; Dr. Shyam Kishore. G; Dr. Jongoni Srikanth; Ms. Shobana D; Dr M. Jeyaselvi; Dr. Vijay Vasanth A; Ms. D. Linett Sophia; Dr. B. Mahesh; Dr. M. Shanmuganathan; and Dr. Jeba Johannah J.
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 present invention is directed to an AI-based low-power architecture for future IoT Devices. The proposed framework mitigates the main issues of existing IOT systems such as, high energy consumption, ineffective computation, the overhead of communication via incorporating a unified framework of lightweight AI, reinforcement learning, and energy harvesting. The architecture features a small-footprint AI engine designed to run efficiently in embedded inferencing applications that harness on-device intelligence from model compression, and pruning to event-driven, neuromorphic-like processing. This reduces the computational burden over the inference and also helps for good inference. A power management system, based on reinforcement learning, acts as a smart device that learns usage patterns and adjusts energy between sensing, computation and communication on the fly. Power harvesting modules can be connected to further complement the architecture, making battery-limited devices operate sustainably. Furthermore, the proposed methodology also facilitates hybrid computation, allowing to offload workloads in a selective way to fog or edge nodes on the basis of the AI-based decision models for managing latency, energy, and bandwidth. Communication energy is saved through an adaptive protocol, powered by context-aware AI, to minimize duplicate transmissions and favor important transmissions. Federated learning guarantees both scalable and privacy-preserving model updates without sending raw data, and lightweight cryptography which in turn offers security for communication in low-power design limits. The architecture can be deployed across the applications of smart healthcare, industrial IoT, environment monitoring, and 6G based smart cities. This innovation thus offers a fundamentally new, AI-centric, energy-efficient IoT framework that guarantees autonomous, secure and scalable operation, opening the door to sustainable future IoT ecosystems."
Disclaimer: Curated by HT Syndication.
