MUMBAI, India, Oct. 12 -- Intellectual Property India has published a patent application (202641118531 A) filed by Madhankumar C; Mrs. M. Shyamala - Hindusthan Institute Of Technology; Mrs. Akilandeswari N - Jct College Of Engineering And Technology; Mrs. V. Karpagam - J. J. College Of Engineering And Technology; Dr. V. Vijayakumar - V. S. B. Engineering College; Ms. A. Ramyasri - Erode Sengunthar Engineering College; Mrs. S. Subasree - Care College Of Engineering; Dr. M. Gokilavani - Dayananda Sagar University; M. S. Vinu - Jct College Of Engineering And Technology; Ms Sruthi S. Madhavan - Nehru Institute Of Engineering And Technology; Dr. M. Maheswaran - Nehru Institute Of Engineering And Technology; Dr. Palani Kannan - Info Institute Of Engineering; and Ms. R. Devi - R. M. K. Engineering College on October 04, 2026, for Neuromorphic Edge Computing Architecture For Ultra-Low Latency Autonomous Iot Intelligence.
Inventors include Mrs. M. Shyamala - Hindusthan Institute Of Technology; Mrs. Akilandeswari N - Jct College Of Engineering And Technology; Mrs. V. Karpagam - J. J. College Of Engineering And Technology; Dr. V. Vijayakumar - V. S. B. Engineering College; Ms. A. Ramyasri - Erode Sengunthar Engineering College; Mrs. S. Subasree - Care College Of Engineering; Dr. M. Gokilavani - Dayananda Sagar University; M. S. Vinu - Jct College Of Engineering And Technology; Ms Sruthi S. Madhavan - Nehru Institute Of Engineering And Technology; Dr. M. Maheswaran - Nehru Institute Of Engineering And Technology; Dr. Palani Kannan - Info Institute Of Engineering; and Ms. R. Devi - R. M. K. Engineering College.
The application for the patent was published on October 09, 2026, under issue no. 41/2026.
Abstract: ABSTRACT Neuromorphic Edge Computing Architecture for Ultra-Low Latency Autonomous IoT Intelligence The present invention relates to a neuromorphic edge computing architecture for providing ultra-low-latency, energy-efficient, and autonomous intelligence in Internet of Things (IoT) environments. The proposed system integrates distributed IoT sensor nodes, neuromorphic processing units, event-driven sensing, spiking neural networks, edge intelligence, adaptive learning, secure communication, and an autonomous decision-making module within a unified architecture. The system is configured to process sensor information close to the point of data generation, thereby reducing dependency on centralized cloud infrastructure and minimizing communication latency, network bandwidth utilization, and energy consumption. The IoT sensor layer acquires real-time information from heterogeneous sensors including environmental, motion, vibration, temperature, pressure, acoustic, image, industrial, and other application-specific sensing devices. Instead of continuously transmitting raw sensor streams, an event-driven processing mechanism identifies meaningful changes or events and generates corresponding event representations. The neuromorphic edge-processing unit processes such event-driven information using spiking neural network models and neuromorphic computing principles to perform real-time pattern recognition, anomaly detection, classification, prediction, and decision-making. An adaptive learning module enables the neuromorphic edge nodes to update or adjust computational parameters according to changing environmental and operational conditions. The system may dynamically modify processing thresholds, synaptic weights, event priorities, and resource-allocation parameters based on observed sensor patterns and system feedback. A resource-management module allocates processing, memory, communication, and energy resources according to computational requirements and event priority, thereby supporting continuous autonomous operation on resource-constrained IoT devices. The architecture further incorporates a distributed intelligence mechanism in which multiple edge nodes can exchange selected features, event information, or decisions rather than transmitting continuous raw sensor data. A secure communication module provides authenticated and encrypted communication between IoT devices, neuromorphic edge nodes, and authorized management systems. Local decision-making allows the system to continue performing configured intelligence functions even when connectivity to a remote cloud or central server is unavailable. The proposed architecture can perform real-time applications including anomaly detection, predictive maintenance, environmental monitoring, smart transportation, industrial automation, healthcare monitoring, energy management, intelligent robotics, and autonomous IoT control. An explainability or decision-trace mechanism may additionally record relevant sensor events, temporal patterns, and decision factors associated with an autonomous output, enabling authorized users to review the basis of system decisions. Accordingly, the invention provides a low-latency, energy-efficient, adaptive, secure, and autonomous IoT intelligence platform by combining event-driven sensing, neuromorphic computing, spiking neural networks, edge processing, adaptive learning, and distributed decision-making. The architecture reduces unnecessary data transmission and centralized processing while enabling rapid local intelligence for next-generation IoT applications.
Disclaimer: Curated by HT Syndication.
