Unifying Edge and Cloud Computing: A Framework for Distributed AI and Real-Time Processing

Abstract

Edge computing and cloud computing are two popular computing paradigms that have distinct application scenarios and are often regarded as mutually exclusive. However, they have their own advantages and are often used together to support innovative applications like intelligent transportation, smart homes, and video analytics in smart cities. As edge nodes operate remotely with limited computing resources, different from cloud computing, applying AI to distributed environments facilitates real-time processing and the exchange of information. In addition, the way that unifies edge and cloud computing makes it preferable for engineering robot teams.
 
In this paper, we propose a cloud-edge computing migration framework for merging AI at the edge with the real-time processing power of the cloud, a different distributed AI framework, necessary components and algorithms, and the primary drivers of the aforementioned solution from a systematic point of view. At the same time, we collect the latest progress and clarify why the mentioned environment has a huge impact on applications such as collaborative robotics and self-driving vehicles. Cloud computing is at the other end of the continuum from edge computing, operating through the web with its resources, applications, and services. Depending on general-purpose processing, highly available services can be activated in seconds with the load of a button.