IJAEMS
IJAEMS
The Artificial intelligence is gradually shifting from narrow applications to more sophisticated solutions capable of providing real-time adaptive decision-making across various operational environments. In line with the current trend, this paper offers a unified edge-cloud intelligence framework that is aimed at incorporating the following areas in one cross-domain system: public safety video analytics, personalized marketing, consumer sentiment monitoring, and aerospace navigation. With the use of edge computing, low-latency data processing and inference will be performed closer to the source, while reinforcement learning algorithms implemented in the cloud environment will ensure continuous model refinement, adaptive intelligence, and big data processing. This collaboration between edge and cloud resources will allow efficient utilization of heterogeneous environments with different computing needs, network connectivity, and data streams. The efficiency of the framework was evaluated using a series of comprehensive experiments in domain-specific scenarios. The outcomes showed that the unified intelligence framework outperformed the traditional single-domain artificial intelligence models by 27% in terms of inference latency, by 15% in terms of prediction accuracy, and by 22% in terms of adaptability. The public safety scenario resulted in near-real-time detection of anomalous events with the precision of more than 94%. The marketing case studies helped to improve the user. This approach allows for smooth collaboration between distributed edge computing systems and centralized cloud-based intelligence, allowing for quicker reaction times and optimized use of resources. This architecture is free from the limitations that are imposed on stand-alone systems with domain-specific knowledge by providing an opportunity to exchange knowledge and adapt learning in different application areas. Moreover, this architecture can accommodate continuous flow of data while remaining accurate in making decisions. These features make it possible for this proposed framework to be used in next generation intelligent systems.