Artificial intelligence-driven edge computing and cloud synchronization system in field of big data analytics, has resource optimization engine that selectively offloads compute tasks to cloud server based on edge-device resource availability, latency constraints and data criticality
	MADHALA R; SHAIK B; SALAMKAR M A; IMMANENI J; BODA V V R
	2025-03-28
专利权人	MADHALA R (MADH-Individual) ; SHAIK B (SHAI-Individual) ; SALAMKAR M A (SALA-Individual) ; IMMANENI J (IMMA-Individual) ; BODA V V R (BODA-Individual)
申请日期	2025-03-28
专利号	IN202541030571-A
成果简介	NOVELTY - The artificial intelligence (AI)-driven edge computing and cloud synchronization system (100) has an edge computing module (104) for processing sensor data locally via an adaptive machine learning (ML) model to generate real-time insights. A cloud synchronization module (108) comprises a two-way communication channel between an edge device (102) and a remote cloud server and is configured to: transmit compressed edge-processed data to the cloud server for aggregate analysis, receive updated machine learning parameters from the cloud server based on the aggregate analysis and dynamically retrain the adaptive machine learning model on the edge device using the updated parameters. A resource optimization engine (110) selectively offloads compute tasks to the cloud server based on edge-device resource availability, latency constraints and data criticality. USE - AI-driven edge computing and cloud synchronization system in field of big data analytics. Can also be used for devices integrating AI, localized edge computing and automated synchronization with cloud infrastructure. ADVANTAGE - The method t enables on-device incremental learning, allows edge AI models to adapt to local data patterns without requiring full cloud retraining, provides cross-validated against cloud predictions, triggers retraining when confidence thresholds are breached, contributes to global model training in the cloud without exposing raw data, preserves privacy while improving model generalizability, allocates compute tasks based on real-time edge-device metrics and network conditions, achieves selective offloading, reduces edge-device power consumption by 30-50% and inference latency by 20-30%, extends battery life in IoT deployments, avoids adversarial tampering and cloud dependent delays during peak shopping hours, trains robust global models without centralized data storage and optimizes energy distribution in real time. DETAILED DESCRIPTION - An INDEPENDENT CLAIM is also included for a method for AI-powered edge computing and cloud synchronization. DESCRIPTION OF DRAWING(S) - The drawing shows a block diagram of the AI-driven edge computing and cloud synchronization system. 100AI-driven edge computing and cloud synchronization system 102Edge device 104Edge computing module 106Network 108Cloud synchronization module 110Resource optimization engine
IPC 分类号	G06F-018/214 ; G06F-009/50 ; G06N-020/00 ; G06N-003/045 ; H04L-067/10
国家	印度
专业领域	信息技术
语种	英语
成果类型	专利
文献类型	科技成果
条目标识符	http://119.78.100.226:8889/handle/3KE4DYBR/13428
专题	中国科学院新疆生态与地理研究所
作者单位	1.MADHALA R (MADH-Individual) 2.SHAIK B (SHAI-Individual) 3.SALAMKAR M A (SALA-Individual) 4.IMMANENI J (IMMA-Individual) 5.BODA V V R (BODA-Individual)
推荐引用方式 GB/T 7714	MADHALA R,SHAIK B,SALAMKAR M A,et al. Artificial intelligence-driven edge computing and cloud synchronization system in field of big data analytics, has resource optimization engine that selectively offloads compute tasks to cloud server based on edge-device resource availability, latency constraints and data criticality. IN202541030571-A[P]. 2025.

条目包含的文件
条目无相关文件。

除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。