The Investigation and Challenges of Advanced Applications Using Artificial IoT on Edge Computing

Abstract

The rapid development of Internet-of-Things (AIoT) systems that assess and respond intelligently to environmental stimuli without human participation has been substantially facilitated by the combined integration of AI and the IoT. However, the volume, velocity, and validity of data and catastrophic transmission latency make it difficult or impossible to process huge volumes of data on the cloud. Edge computing is a viable solution for these pressing problems. In the beginning it defines certain broad terms like "Internet of Things," "artificial intelligence," and "edge computing." Informed by these ideas, it investigates the broad architecture of AIoT, provides a real-world AIoT example to show how AI can be implemented in the real world, and analyzes promising AIoT use cases. In the other side, a general look at the state of the art in AI model inference and training at the network's edge have been showed. At the end, the remaining problems and potential future developments in this field are discussed in details. AIoT will also be accompanied with concerns such as security, data privacy, and ethical difficulties. The purpose of this research is to conduct an in-depth analysis of edge computing and the impact that AI has on the Internet of Things, a machine learning program determine how much data can be sent to the cloud with little loss of quality, AIoT architecture and provides a practical AIoT example to show how AI can be implemented in real-world. The analyzing results shows the uses, outcomes, and difficulties associated with AIoT and edge computing.