Integrating Cyber-Physical Systems, IoT, and Cloud Computing for Intelligent and Scalable Applications

Abstract

The integration of Cyber-Physical Systems (CPS), the Internet of Things (IoT), and Cloud Computing has emerged as a key enabler for intelligent, scalable applications; however, challenges related to latency, scalability, energy efficiency, resource utilisation, and learning accuracy remain unresolved. This study aims to design and experimentally evaluate an end-to-end CPS–IoT–Cloud architecture that addresses these challenges in a unified framework. The proposed methodology integrates sensors and actuators, edge and gateway devices, cloud-based analytics, and intelligent applications, and is evaluated under varying device densities, workloads, and training data volumes. Experimental results show that the integrated architecture reduces end-to-end latency by approximately 38–39% compared to a CPS-only system when scaling up to 1,000 devices, while improving throughput by up to 40% over baseline architectures. Energy consumption is reduced by an average of 22% across sensors, edge nodes, and gateways. Cloud resource analysis shows efficient scalability, with CPU utilisation reaching about 92% and memory utilisation remaining below 80% at peak workload. In addition, cloud-based intelligence achieves prediction accuracy of up to 98–99%, outperforming edge-based intelligence by 10–11 percentage points at large training data sizes. The novelty of this work lies in its holistic, experimentally validated evaluation across all CPS–IoT–Cloud layers using consistent metrics. The results confirm that coordinated CPS–IoT–Cloud integration provides a practical and scalable solution for intelligent applications such as smart cities and industrial automation.