Hologram AR tech provider WiMi Hologram Cloud has integrated AI and Augmented Reality together onto an edge computing platform. The company is in the process of launching a MEC-based intelligent AR system platform.

WiMi’s platform relies on AR terminal devices to collect images, audio, and video information. It uploads them to the AI analysis and processing system deployed on the edge cloud through the 5G network and quickly studies and processes them through AI to provide supplementary support information for users and display them in the AR terminal devices. WiMi integrates the MEC edge cloud with the 5G communication technology network and combines it with emerging intelligent applications, gradually forming a multidimensional service system of “5G, MEC, and intelligent applications”. As an overall solution for MEC in the 5G era, it can improve work efficiency and service level and has broad application prospects in industries such as industrial internet, smart city, innovative medical, and intelligent security.

WiMi’s MEC-based Intelligent AR System Platform is divided into a two-tier architecture. The Edge cloud platform is a kind of cloud platform using the traditional cloud computing IaaS, PaaS, and SaaS layered architecture. The edge infrastructure layer adopts a lightweight architecture, providing computing, storage, and network acceleration capabilities.

The edge PaaS layer introduces video trans-coding, AR rendering, and AI besides traditional functions and services. The 5G edge UPF is presented in the edge cloud to provide triage capabilities based on IP, DNS, and device ID so that users can configure data access policies to enable data flow to the central cloud or remain in the local processing.

The central cloud provides computation and management functions. The computation function adopts the three-tier architecture of cloud computing, providing advanced processing capabilities so as to handle the demanding computing situations that the edge cloud may not tackle due to its lightweight architecture. Capabilities such as AI training and high-level AR rendering provided by the PaaS layer can be handled by the central cloud due to the high computing power requirements. The management function of the central cloud provides edge DC management, IaaS management, PaaS management, and application management that can coordinate and manage multiple edge cloud nodes and their infrastructure, PaaS, and applications within the system.

The edge cloud and the central cloud work together through a linkage mechanism. The edge collection devices of the platform will collect a large amount of data, which will all be uploaded to the edge cloud for processing. The edge cloud can deal with the processing and storage of AR and AI within a specific range. Combined with AI’s intelligent analysis and processing technology, it is oriented toward business scenarios such as AR image rendering, integration, video monitoring, face recognition, etc. With low latency, large bandwidth, and quick response characteristics, it can make up for the current problems of high latency and poor user experience and realise local analysis, fast processing, and real-time response. Most of the data collected by terminal devices can be used as data sources for big data processing, and the data can be uploaded to the central cloud after processing or directly uploaded to the central cloud by the collection devices.