After the concept of 5G broke out in 2019, the concept of edge computing has also been rapidly popularized. The share prices of listed companies related to edge computing have risen and stopped for many days, and the reading volume of articles about edge computing has increased several times.What are the functions and application scenarios of products under this concept? Next, we will discuss three issues: what is edge computing, edge computing and cloud computing, and edge computing application scenarios.
The term edge computing, which began to appear in 2011, has become a hot investment spot for giants, referring to the processing, analysis and storage of data closer to the location of data generation, so as to achieve fast and near-real-time analysis and response.
Popularly speaking, edge computing is essentially a service, similar to cloud computing and big data services, but this service is very close to users, providing computing, storage, network and application services closer to users.
The problems that edge computing focuses on are high latency, network instability and low bandwidth in the traditional cloud computing model.In the traditional cloud computing mode, due to poor network conditions and bandwidth limitations of cloud servers, it is inevitable to be affected by high latency and network instability, thereby affecting system efficiency, but by sinking part or all of the processing procedures to the edge side close to the user or data source, the impact on the cloud center mode can be greatly reduced.
Since the development of cloud computing technology, some companies have consolidated their operations by centralizing data storage and computing in the cloud. However, in the context of the interconnection of all things, the number of terminal devices is growing exponentially, hundreds of millions of devices bring massive heterogeneous data, new scenarios and new needs are frequent, so that the traditional cloud computing model is facing enormous challenges.
Take Boeing 787 as an example, each round trip can generate terabyte-level data, and the United States collects 3.6 million flight records every month. Monitor 25000 engines in all aircraft, each generating 588 GB of data a day. Such a level of data, if uploaded to the cloud computing server, both for computing power and bandwidth, have put forward stringent requirements.
For another example, wind turbines are equipped with a variety of sensors to measure wind speed, pitch, oil temperature, etc., which are measured every few milliseconds to detect the wear of blades, gearboxes, frequency converters, etc. A wind farm with 500 wind turbines will generate 2 PB of data a year.
If such petabyte-level data is uploaded to the cloud computing center in real time and makes decisions, it puts forward stringent requirements in terms of computing power and bandwidth, not to mention the problem of immediate response due to delay.
Challenges facing cloud computing, the main factors include:
Delay. More industries are implementing applications that require rapid analysis and response. Cloud computing alone cannot keep pace with these needs because the network distance of data sources creates latency, resulting in inefficiencies, time delays, and poor customer experience.
Bandwidth. Cloud computing latency can be addressed by increasing transmission bandwidth or more processing power. However, as companies continue to increase the number of edge devices in their networks and the amount of data they generate, the cost of delivering data to the cloud will reach unrealistic levels.
Security and privacy. Sensitive data such as private medical records and business data are transmitted to cloud data centers through the Internet, which greatly increases the risk of data interception. In terms of security, some governments or customers may need to keep data in the jurisdiction where it was created.In healthcare, for example, there may even be local or regional restrictions on the storage or transfer of personal data.
Connectivity. The lack of a persistent Internet connection can hinder cloud computing, but a variety of network connectivity options can support edge-to-cloud computing. For example, 5G provides high-bandwidth and low-latency connectivity for fast data transfer and edge service delivery.
Artificial intelligence. Because of the need for near-real-time, actionable data, companies need AI closer to the data source to speed up processing and harness the potential of previously untapped data.
By 2025, it is estimated that 75% of all data will be generated outside of the central data centers where most of it is now generated.Faced with this scenario, edge computing shows its advantages. Because it is deployed near the device side, it can make real-time feedback decisions through the algorithm, and can filter most of the data, effectively reducing the load of the cloud, making massive connections and massive data processing possible.
Benefits of edge computing:
Increase speed and decrease latency. Moving data processing and analytics to the edge can help speed up system response, thereby accelerating transaction processing and enhancing the experience, which is critical in near-real-time applications, such as autonomous vehicles.
Improve network traffic management. Minimizing the amount of data sent over the network to the cloud reduces the bandwidth and cost of transmitting and storing large amounts of data.
Higher reliability. The amount of data that a network can transmit at a single time is limited. For locations with poor network connectivity, the ability to store and process data at the edge improves reliability in the event of a cloud outage.
Enhanced security. Properly implemented, edge computing solutions increase data security by restricting the flow of data across the network.
While edge computing offers unprecedented opportunities to unlock the value of data, the cloud remains a necessity for central databases and processing centers. In the future architecture of the Internet of Things, edge computing will be a supplement to cloud computing, and there is no question of who will completely replace whom. Different scenarios bring different requirements, and different requirements require different network architectures.The diversification of scenarios is a reality, so the flexibility of network architecture is also an inevitable choice.
3. Application scenarios of edge computing
Edge computing is an important supplement to cloud computing. Its basic feature is to bring computing power closer to users, that is, sites are widely distributed and edge nodes are connected by wide area networks. There are mainly the following 13 application scenarios:
Medical treatment-remote surgery, remote monitoring, etc.
Telemedicine, through the Internet communication technology to assist the completion of health care process, to achieve "long-distance treatment". Such as remote pathological diagnosis, remote medical imaging diagnosis, remote consultation, remote surgery, remote monitoring and so on. Telemedicine can greatly facilitate medical treatment, reduce the number of patients in the hospital, and alleviate the difficulty of seeing a doctor.Patients in remote areas can receive joint treatment from doctors in local clinics and third-class hospitals and enjoy high-quality medical resources.
Transportation-Smart Driving Car
With the popularization and application of emerging technologies such as 5G, Internet of Things and edge computing, the automotive industry is facing a huge transformation, and intelligent driving (including assisted driving, automatic driving and driverless driving) has gradually become a reality.
Finance-Unmanned Banking
The use of unmanned banks demonstrates the latest financial intelligent technology achievements such as biometrics, speech recognition, data mining, and integrates the current hot technology elements such as robots, VR, AR, face recognition, voice navigation, holographic projection and so on.
Industry-Industrial Manufacturing
Intelligent manufacturing is the inevitable trend of the development of advanced manufacturing technology in the future. As a very important part of the intelligent manufacturing industry, the application scenarios of edge computing in the intelligent manufacturing industry are also growing.
Education-The Classroom of the Future
The future classroom is a high-quality solution to realize interactive learning, flipped classroom and interactive classroom according to the needs of teaching mode reform and innovation. It uses advanced information equipment to realize the innovation of double-screen teaching, interactive teaching, group discussion teaching, classroom direct recording and broadcasting, remote synchronization and other teaching modes, so as to provide guarantee for students'dominant position.
Logistics-Smart Logistics
By deploying edge computing infrastructure and supporting the logistics platform to achieve edge computing capabilities, the perception interconnection among vehicles, drivers/transport owners/cargo owners, cargo and road infrastructure can be achieved.It is of positive significance to improve the efficiency of transportation management, standardize the driving behavior of drivers, ensure the quality of goods, the safety of vehicle operation, improve the efficiency of traffic operation and reduce pollution emissions.
City-smart parking, street light control, etc.
Intelligent parking is supported by edge computing and Internet of Things (loT) related technologies. When a vehicle is detected to enter the garage, it is immediately fed back to edge computing, which calculates and triggers the indoor navigation system to provide parking reservation and route planning for consumers. Really solve the customer's parking difficulties and troubles.
Power-wireless communication control and acquisition
Smart grid wireless communication application scenarios can be generally divided into two categories: control and acquisition. Among them, the control category includes intelligent distributed distribution automation, power load demand side response, distributed energy regulation and control, etc. The acquisition category mainly includes advanced metering, smart grid and large video applications.
Security-Safe City
In the face recognition business of security and unmanned shopping malls and supermarkets, the edge node is used to deploy Al face recognition service, compare the face database locally, return the calculation results, and then upload the necessary information to the central database for storage and multi-place information synchronization.Using the edge to deploy Al face recognition service, on the one hand, it can quickly return the results and reduce the service delay; On the other hand, it reduces the transmission of unnecessary images, videos and other large amounts of data in the backbone network, and only transmits the necessary characteristic information, which reduces the bandwidth cost.
Home-Smart Home
In order to meet the requirements of real-time monitoring, timely response and reliable operation of smart home system, the importance of edge computing is becoming more and more prominent, and the linkage between devices can be realized through edge computing in LAN.Through the integration of edge computing, Internet of Things and related edge computing protocols, the intelligent control of various terminal devices is realized, and the lag problem caused by network delay in the past is solved.
Buildings-smart buildings, smart access control, etc.
For smart building construction, edge computing will play a huge role in building intelligent office optimization and intelligent security optimization. Through edge computing deployment, the management department can timely handle demand monitoring, key area access control monitoring, video surveillance, air conditioning management and fire management, and realize real-time intelligent optimization of the building.The most obvious is in the optimization of building intelligent power distribution, smart meters and upstream supply equipment work together to predict regional electricity demand, optimize the utilization of distribution resources, reduce power outages and waste.
Service industry-unmanned supermarket, unmanned hotel, unmanned restaurant, etc.
By deploying edge computing to compute data locally, robots can achieve environmental awareness, human-computer interaction, decision-making and control, and can solve problems more economically in real time.
Agriculture — — Plant protection UAV, factory seedling, etc.
Intelligent agriculture (or factory agriculture) refers to the modern advanced agricultural production mode which adopts industrial production to achieve intensive, efficient and sustainable development under relatively controllable environmental conditions. Plant protection UAV can not only spray pesticides, but also collect and monitor data. Intelligent greenhouse can intelligently adjust indoor temperature, light, water, fertilizer, gas and many other factors.All these have the requirements of low delay, high bandwidth, high stability and localization.
The next few years will be a critical period for the development of edge computing. There is an urgent need for all parties in the industry to promote the integration and development of edge computing with 5G, Internet of Things, artificial intelligence and other technologies, and accelerate the landing of new technologies and industries.
