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

Cloud computing is no longer sufficient to instantly process and analyze vast amounts of data generated (or about to be generated) by Internet of things devices, connected cars, and other digital platforms. For example, a vehicle with self-driving is essentially a high-performance computer on the wheel that collects data through multiple sensors. In order for these vehicles to run safely and reliably, they need to respond immediately to their surroundings. Any lag in processing speed can be fatal.

Although most of the data processing for connected devices is now in the cloud, it can take a few seconds to send data back and forth on a central server. By 2020, it is estimated that smart cars will generate 1.5 GB of data per day. As more devices connect to the Internet and generate data, cloud computing may not be able to handle it all-especially in some uses that require faster response times.

Edge computing can provide faster data processing for self-driving cars, but how does it work?

Edge computing enables connected devices to process data closer to where they were created-or "edges". This can be within the device itself (that is, the sensor) or near the device, providing an alternative to sending data to a centralized cloud for processing. Some of the biggest technology companies, including Amazon (AMZN), Microsoft (MSFT) and Google (GOOG), are exploring edge computing, which could trigger the next big computing race.

The shift to marginal computing

In our data-intensive future, faster and more reliable data processing will become critical as billions of devices connect to the Internet. In recent years, the integration and centralization of cloud computing has proved to be cost-effective and flexible, but the rise of the Internet of things and mobile computing has put pressure on network bandwidth.

In the end, not all smart devices need to use cloud computing to run. In some cases, this repeated data transmission can and should be avoided. This is the opportunity for marginal computing to give full play to its talents. According to the CB Insights market size assessment, the global marginal computing market is expected to reach US $6.72 billion by 2022.

Although it is an emerging area, edge computing may be more efficient in some areas of cloud computing operations. Edge computing brings data processing closer to where it was created (for example, motors, pumps, generators, or other sensors), reducing the need to transfer data back and forth between clouds.

According to research firm IDC, edge computing is described as a "mesh network of micro data centers that can process or store critical data locally and push all received data to a central data center or cloud repository, covering an area of less than 100sq ft".

For example, a train may contain a sensor that immediately provides the status of its engine. In this case, whether the data is running on the train or in the cloud, the sensor data does not need to go to the data center to see if something is affecting the operation of the transmitter.

For example, a train may contain a sensor that immediately provides the status of its engine. In this case, whether it is running data on the train or

Localizing data processing and storage can reduce the burden on computing networks. When less data is sent to the cloud, the likelihood of latency (the data processing delay caused by the interaction between the cloud and the Internet of things) is reduced. This also takes more responsibility for the underlying edge computing technology of the hardware, which includes sensors for collecting data and CPU or GPU for processing data within connected devices.

Why Intelligent Manufacturing Industry is optimistic about Edge Computing

Intelligent manufacturing can be inspired by the large number of sensors used in modern factories. Reducing the latency of edge computing can lead to faster and more sensitive changes in manufacturing workflows, allowing insights and operations to be applied in real time. This may include turning off the machine before it overheats. Factories can use two robots, equipped with sensors and connected to edge devices, to perform the same task. Edge devices can run machine learning models to predict whether one of the robots will fail (as shown in the figure).

If the edge device determines that the robot may fail, an action is triggered to stop or slow it down. This will allow the factory to assess potential failures in real time. If robots can process data on their own, they may become more self-sufficient and responsive. Edge computing should allow greater and faster insight to be generated from big data, and more machine learning can be applied to operations.

The ultimate goal is to take advantage of the undeveloped value of the large amount of data being created to prevent safety hazards and reduce disruptions on the factory floor. From wearable devices to smart factories to smart driving, the development of Internet of things devices is getting wider and wider. Edge computing can provide another efficient solution to develop software and hardware capabilities around device functions, breaking the limitations of traditional architectures for centralized cloud data center data management, processing, and storage.

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.