Shulou(Shulou.com)06/02 Report--

Edge computing makes the park full of "wisdom"

Edge computing refers to the side close to the object or data source, using network, computing, storage, application core capabilities as one of the open platform to provide the nearest end services. Its application is launched on the edge side, which produces faster network service response and meets the basic needs of the industry in real-time business, application intelligence, security and privacy protection. The traditional centralized data center in the cloud is far away from the terminal equipment and users, the data round-trip delay is large, network congestion and other problems are more obvious, the emergence of edge computing perfectly solves these problems.

Since 2018, the trend of edge computing is surging, and it is on a par with cloud computing. Recently, various public cloud vendors have put edge computing in an important position, and edge computing in the industry and media has been mentioned more frequently than cloud computing. From various aspects, edge computing will process more data than the cloud within a year, and will become the main computing and storage node. In addition to the traditional data on the cloud, data analysis, remote monitoring, etc., one of the core competitiveness of edge computing is the ability of local computing and analysis, especially the comprehensive use of a variety of AI and big data technologies, that is, intelligent edge computing.

As more and more manufacturers attach importance to and enter this field, intelligent edge computing is expected to become the hottest area in the industry in 2019. Industry insiders predict that edge computing may be the next tuyere after AI.

Gartner announced the top ten technology trends in 2019, among which edge computing is among them. Gartner sees cloud-to-edge (Cloud to the Edge) as an important trend in the future of technology, and Gartner predicts that 10% of enterprise data will be generated outside the cloud and data center in 2017 and will exceed 50% by 2022. The rapid development of edge computing is mainly due to the following factors:

Low latency: in order to meet the requirements of low latency, a solution needs to be built at the "edge" closest to the business site to reduce business processing latency; massive data: the explosive growth of marginal data in the era of the Internet of things, it is difficult to return directly to the cloud and the cost is high, and the data is analyzed and filtered locally to save network bandwidth. Privacy security: data involves the security of enterprise production and business activities, and deals with enterprise confidential information and personal privacy at the edge; local autonomy: does not rely on the offline processing capability and self-recovery ability of the cloud

In recent years, the maturity of AI technology has become another force promoting edge computing. As the computing and analysis capabilities of these edge devices become stronger and stronger, more and more manufacturers begin to bring machine learning and even deep learning capabilities into the devices, so that today's edge devices can also do what the cloud can do. the application is also getting wider and wider. In addition to AI technology, including AI chip, GPU, network and AI development tools, frameworks and analysis platforms specially built for edge devices have also greatly promoted the development of edge computing.

Edge computing can be widely used in smart parks, industrial manufacturing, merchant supermarkets, vehicle networking and other fields. For example, automatic driving, drones, augmented reality (AR) / virtual reality (VR) and robots, these new applications place special emphasis on real-time image analysis and identification processing capabilities, which require low latency and high bandwidth of the network, and need to respond in tens of milliseconds or even microseconds, but it takes at least hundreds of milliseconds to transmit to and from the cloud through the Internet. Therefore, this type of application is very suitable to adopt the edge computing architecture of "edge cloud collaboration". Through the intelligent pre-analysis and deep learning of the data on the edge side, cloud model training and further big data processing, we can digest the massive data locally and avoid the bandwidth waste and delay caused by the large amount of data return.

Huawei Cloud Intelligent Edge platform IEF services link edge and cloud data to meet customers' demands for remote control, data processing, analysis, decision-making and intelligence of edge computing resources. At the same time, it provides unified device / application monitoring, log collection and other operation and maintenance capabilities in the cloud, providing enterprises with complete edge and cloud collaborative edge computing solutions to help customers easily build smart edges.

Since its launch in April 2018, IEF service has continued to explore in many areas. Among them, Debon Express and Huawei Cloud join hands to explore the construction of a new intelligent logistics park, relying on the edge cloud collaborative architecture based on IEF services to build a smart logistics park on the cloud. This time, the editor will focus on the edge cloud collaborative solution and its inherent value in the wisdom park scenario.

At present, the current situation of park monitoring is mainly focused on "watching" and "saving", and the ability of analysis and situation awareness of surveillance video is weak. The edge cloud collaborative solution based on IEF service realizes the intelligent upgrade from "civil air defense" to "technical defense", which can help customers improve the operation efficiency of the park and improve the resident experience of the park:

1. Low latency: fast local processing of camera code stream to upload pictures to reduce the delay of face recognition

2. Intelligence: intelligent analysis of surveillance video, real-time perception of abnormal events such as intrusion and large flow of people, so as to reduce the labor cost of the park.

3. Convenient management: edge application lifecycle management can be upgraded seamlessly.

4. Model self-learning: automatic training, good algorithm expansibility, and complete self-learning closed loop (from sample collection, training to reasoning)

5. Cost saving: it can take advantage of the existing IPC camera in the old park and change it into an intelligent camera through edge cloud cooperation to save cost.

So from a technical point of view, how does IEF, the intelligent edge platform, endow the smart park with "wisdom"?

First of all, all the cameras in the campus need to be connected to the edge server, and various cloud video analysis algorithms can be sunk to the edge server through IEF services. The video stream captured by the camera can be directly injected into the video analysis algorithm located in the edge server, and the edge video analysis algorithm can directly process the video stream without uploading the video stream data to the public cloud. In this way, it can not only reduce the bandwidth cost, but also improve the real-time processing.

In addition, this scheme is different from the traditional smart camera scheme, the video analysis algorithm in the traditional smart camera is fixed, while the edge cloud service can enable the algorithm flexibly and has strong expansibility. Really enhance the "soft" strength of the camera!

To learn more about edge computing, welcome to Huawei Cloud Institute (https://edu.huaweicloud.com/courses/)

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.