MUMBAI, India, Aug. 17 -- Intellectual Property India has published a patent application (202641096180 A) filed by Dr. Vemuri Venkata Phani Babu; Yaragani Ashok Kumar; Dr. Murthy Ravaleedhar Reddy; Dr. C. Venkataswamy; Dr. K. Rasadurai; Dr. Radha Seelaboyina; Mr. R K Arunkumar; Mr. Pranjal Kaser; Dr. Kishore Kumar M; Mrs. Swaroopa Rani B; Palla Sravani; and M K. Kirubakaran on August 09, 2026, for System And Method For Context-Aware Adaptive Artificial Intelligence Inference On Edge Devices.
Inventors include Dr. Vemuri Venkata Phani Babu; Yaragani Ashok Kumar; Dr. Murthy Ravaleedhar Reddy; Dr. C. Venkataswamy; Dr. K. Rasadurai; Dr. Radha Seelaboyina; Mr. R K Arunkumar; Mr. Pranjal Kaser; Dr. Kishore Kumar M; Mrs. Swaroopa Rani B; Palla Sravani; and M K. Kirubakaran.
The application for the patent was published on August 14, 2026, under issue no. 33/2026.
Abstract: The present invention discloses a system and method for context-aware adaptive artificial intelligence inference on edge devices to enable efficient, low-latency, and resource-aware execution of AI models in dynamic computing environments. The invention comprises a Context Acquisition Module for collecting real-time operational information, a Context Management and Evaluation Engine for generating contextual profiles, an Adaptive Model Management Module for selecting optimized AI models, a Decision and Inference Optimization Engine for determining suitable inference strategies, a Dynamic Resource Allocation Module for distributing computational workloads, an Edge Inference Execution Module for performing local AI inference, an Energy Management Module for reducing power consumption, a Security and Privacy Module for protecting sensitive data, a Communication Management Module for managing local and remote processing decisions, and a Continuous Feedback Learning Module for improving inference performance over time. The system continuously monitors processor utilization, memory availability, battery status, network quality, environmental conditions, workload characteristics, and application requirements to dynamically optimize inference execution. Based on the evaluated context, the system automatically selects appropriate AI models, adjusts execution parameters, allocates computational resources, and performs secure local inference while minimizing cloud dependency. The invention enhances inference accuracy, reduces execution latency, improves computational efficiency, conserves energy, strengthens data privacy, and increases system reliability across heterogeneous edge computing platforms. The disclosed framework supports deployment in Internet of Things devices, smart healthcare systems, autonomous vehicles, industrial automation, intelligent surveillance, wearable devices, robotics, precision agriculture, and smart city infrastructures. By integrating context awareness, adaptive optimization, intelligent resource management, and continuous learning, the invention provides a scalable, secure, and high-performance edge intelligence solution for next-generation artificial intelligence applications.
Disclaimer: Curated by HT Syndication.
