Contributed by Zhang Guilin

According  to The Internet of Things, written by Michael Miller, the author suggests imagining that your car in the future knows who is driving it and adjusts the driving pattern, music, and air conditioning based on your preferences. Imagine that in a factory, each machine in every workshop sends feedback information to help improve the efficiency of production lines; imagine that a city can automatically manage its public lighting, public services, and road maintenance based on real-time conditions and needs.

With fast development of the Internet of Things (IoT), cloud computing, Big Data, and Artificial Intelligence (AI), scenes that used to only be visualized in science fiction films now have become realities. With these emerging technologies, various industries have indeed started the journey of digital transformation, including manufacturing, energy, healthcare, and transportation. As the fourth technical revolution takes place globally, we are stepping into the next era of intelligence.

However, we must proceed with caution.

A wealth of data brings huge opportunities to us but also creates many challenges. Intelligentization helps to enable intelligent decision-making and operations based on intelligent analysis of data. In the era of intelligence with connectivity of everything, exponential amounts of data will come from countless sensors and terminal devices on the edge side. However, traditional systems cannot effectively use this data.

By 2020, more than 50 billion terminals and devices will be connected to the network. By 2018, 50 percent of IoT networks will be bandwidth-constrained, and 40 percent of data will need to be analyzed, processed, and stored on the network edge.

As new data processing requirements are generated, edge computing is put in the spotlight.

Edge Computing Plays an Important Role for Intelligent Transformation

In recent years, various computing architectures have emerged, such as grid computing, cloud computing, box computing, transparent computing, fog computing, and edge computing. Even personnel in the ICT industry may be confused about these different computing architectures.

What is the significance of edge computing since there are already so many computing architectures?

Let’s examine some industry cases:

In the manufacturing industry, IoT technologies are necessary for building smart factories. As a large number of intelligent terminals and devices are connected to the network through industrial networks, enterprises need to calculate and process increasingly large daily service data. Moreover, a large number of industry scenarios require real-time data processing in milliseconds. Due to bandwidth and capacity limitations of the network, the cloud computing architecture cannot achieve real-time responses.

IoT technologies have also been implemented in the oil and gas industry. Massive numbers of sensors are used for automated collection of production data, which significantly reduces human interaction and achieves smart operations of oil and gas fields. However, if all sensors send data to the cloud, the network will become heavily overloaded. In addition, network connections are not stable in many scenarios within this industry.

For the next example, driverless vehicles need to react in environments where high-speed movement is a factor, so the response time is an extremely important indicator. Presume that a vehicle moves at 65 miles per hour. If the automated braking response time is only a few milliseconds slower, the emergency braking distance of the vehicle requires a few extra feet, which may not be enough time to prevent an accident.

From these cases, it can be found that due to the connections of a large number of devices on the network edge side, such as sensors and intelligent terminals, a large amount of real-time data is generated in the physical world. If all of the data is analyzed and processed in the cloud, high WAN bandwidth costs are generated and real-time responses to services on the edge cannot be achieved. Moreover, edge data is highly sensitive and crucial for many industries. Storing such data in the cloud creates security and privacy risks.

To address this variety of challenges, edge computing emerges as a new computing model and technology system. Edge computing integrates network, computing, storage, and application capabilities to provide edge intelligence services on the network edge, near objects or the source of data.

The objective of edge computing is to solve five major challenges encountered by various industries in realizing digital and intelligent transformation through IoT technologies: Connection, Real-time, Optimization, Smart, and Security (CROSS for short), which are exactly the five values of edge computing.

The entire edge computing system includes four key parts: intelligent devices, intelligent gateways, intelligent systems, and intelligent services. Edge computing bridges the physical world and the virtual world.

It is not a coincidence that edge computing has become essential. With this transition into an era of connectivity of everything, edge computing adds great significance in promoting digital and intelligent transformation of industries.

The essence of digital transformation for industries is to generate data from people, things, the environment, and processes through digitalization, to realize the value flow of data through networking, and to take data as a production factor to create both economic and social values for various industries through intelligence. The ultimate goal of digital transformation of industries is to achieve industry intelligence, which is defined by the following three phases.

Phase 1.0 is business intelligence; phase 2.0 is smart connectivity of everything; and phase 3.0 is end-to-end process collaboration. The three phases have different requirements on the technology side. Phase 1.0 requires cloud computing and Big Data. Phase 2.0 and phase 3.0 require defining new computing models and technology systems on the network edge side to intelligently distribute connections and data analysis closer to the objects or the source of data.

In other words, to achieve industry intelligence, edge computing is a necessary accelerator.

Industry Consortium Is Essential

Edge computing has shown great potential for Augmented Reality (AR), Virtual Reality (VR), drones, Unmanned Aerial Vehicles (UAV), smart manufacturing, oil and gas, healthcare, and Smart City.

As any emerging technology system evolves, so edge computing also encounters its own challenges.

The first challenge is cross-industry collaboration. To achieve intelligence, industries including manufacturing, energy, and utilities need to integrate technologies with other industries, such as machinery, electronics, and ICT. However, Operational Technology (OT) and ICT have traditionally been divided into two separate fields. Edge computing must first enable deep collaboration between OT and IT, and combine industry-specific technologies and knowledge with ICT digital technologies.

The second challenge is industry chain integration. Both OT and ICT have complex technology systems and industry chains. Taking the ICT field as an example, intelligence cannot be separated from collaboration in areas such as sensing, network, computing, storage, and data analysis. The commercialization and implementation of the final solution requires multi-role collaboration among chip manufacturers, hardware product manufacturers, system integrators, and application developers in the industry chain.

The third challenge is technology fragmentation. Every field in the edge technology system involves plenty of technologies: There are over six industrial real-time Ethernet technologies in the industry, more than 40 types of industrial buses, and a variety of public and private cloud platforms. Technology fragmentation poses great challenges to interoperability between systems and exploration of data values and leads to increasing costs.

The fourth challenge is technical uncertainty. Applications of new technologies, such as AI and blockchain, are in the early stages of exploration, so uncertainty risks exist. Early adopters of these technologies want to reduce risks in technology investments and gain advantages from technology applications.

To address these challenges, collaborative efforts throughout the entire ecosystem are needed.

In 2015, edge computing was added to Gartner’s Hype Cycle. Subsequently, industrialization of edge computing increased. Various industrial and commercial organizations have been actively initiating and promoting activities such as research, standardization, industrialization, and commercialization of edge computing.

To represent such growth, in November 2016, the Edge Computing Consortium (ECC) was launched in China by companies including Huawei, Shenyang Institute of Automation (SIA), China Academy of Information and Communications Technology (CAICT), Intel, ARM, and iSoftStone.

The vision of the ECC is to promote advancements in the edge computing industry and to deepen digital transformation of industries. From the very beginning, ECC defined four major positions: Setting up an edge computing industry cooperation platform, promoting open collaboration between OT and ICT industries, incubating the industry’s best application practices, and promoting the healthy and sustainable development of the edge computing industry.

Additionally, ECC planned three major business lines. First, the reference architecture was defined to unify the terms and the architecture and to promote industrial collaboration. Second, a test bed process was defined and the process is promoted as a model. And, last, extensive cooperation among companies is needed to build an industrial ecosystem.

One year after its establishment, ECC has accomplished many achievements.

For ecosystem collaboration, the number of ECC members increased from 62 at its inception to 136 at the end of October 2017, including internationally renowned enterprises such as Mitsubishi, Bosch, Schindler, Schneider, Phoenix, Works Systems, Wind River, Infosys, McAfee, and ABB. During this past year, ECC also established official cooperation relationships with international standards organizations, including IIC, IEC, JTC-1/SC41, IEEE P2413, and IETF.

As for accelerating the industrialization and commercial monetization of edge computing, ECC and industry partners have jointly created more than 11 test beds involving fields such as industrial manufacturing, Smart City, electric power, and transportation. The test beds include those for machine tool IoT, lighting IoT, smart building/water, and intelligent vehicle management system.

On November 7, 2017, ECC co-hosted the Edge Computing Forum (ECF) with Fraunhofer Institute for Open Communication Systems (FOKUS), Europe’s largest research institute for applied science. ECF is the first edge computing summit held in Europe. After the gradual establishment of an ecosystem in China, ECC has taken the first steps towards internationalization.

At the initial inception of the ECC, Swift Liu, the Vice Chairman of ECC and the President of Huawei Network R&D Dept., compared the role of ECC to “the Laval nozzle in a spacecraft engine”. Through continued cooperation within the ecosystem, ECC standardizes the applications of edge computing in some industries such as manufacturing, electric power, and public utilities, and further promotes the concepts in all industries, thereby accelerating digital and intelligent transformation for industries.

The second session of ECC committee is scheduled to be held by the end of November 2017. Let’s stay tuned to see how industries will benefit from the “nozzle” and what new results will be achieved with this collaborative ECC ecosystem.

 

The post Why Has Edge Computing Become the New Hotspot in the Computing Industry? appeared first on Huawei Enterprise Blog.

Source: Huawei Enterprise Blog
—