Airplanes are beginning to integrate data centers in which numerous sensors collect data. This data can be used to aid airlines in maintenance and upgrades, and to facilitate operations and management (O&M), service planning, flight and fuel analysis, navigation services, and recovery. As an elevator becomes intelligent, its running status can be remotely monitored in real time. IoT-connected elevators bring predictive maintenance amongst other benefits. In our world, such the applications of edge computing cannot be counted. This is an age where the impossible becomes reality.

Why has digital transformation surged recently? Because it is inevitable. It significantly boosts industrial automation and meets market requirements for personalized products and services. Most importantly, digital transformation promotes full-lifecycle transformation in products to service operation. This leads to innovations in product services and business models, creating long-term and profound effects on the development of the value and supply chains as well as the market eco-system.

Two Fundamental Technologies of Digital Transformation

It is unquestionable that digital transformation will be an important trend in 2017. People have familiarized themselves with relevant information and ICT matters in specific areas such as digital transformation, cloud computing, and Big Data. On both data centers and the cloud, innovative technologies and solutions continue to emerge. From a general perspective, the global digital revolution is inciting a new wave of industrial transformation. The most obvious difference is the integration into intelligent interconnections, achieving in-depth collaboration and integration of operational technology (OT) and ICT.

Typically, to adjust for digital transformation, the following must be considered:

The relationship between cloud and edge computing—similar to the relationship between the cloud and devices

Integration of OT and ICT

The matching relationship between technical innovations and industry applications

Man-to-thing and thing-to-thing relationships

Omnipresent connectivity and security of connections and devices

To smoothly achieve digital transformation, we must be aware of the omnipresence of intelligent interconnections. Digital transformation has affected numerous areas, such as predictive maintenance in the airline industry, intelligent elevator operation in the public affairs sector, intelligent meter reading in the energy industry, and full process tracing in the logistics industry. All of these industry digital applications show the deep impacts IoT will inevitably trigger in industries such as manufacturing, energy, public affairs, transportation, healthcare, and agricultural sectors. IoT is the foundation of the Made in China 2025 strategy, North America's Industrial Internet, and Europe's Industry 4.0.

The cooperation of cloud computing and edge computing is instrumental in enabling the digital transformation of industries. Cloud computing focuses on Big Data analytics for non-real-time and long-period data, effectively improving periodic maintenance and decision making. Edge computing is for real-time and short-period data, supporting local real-time and intelligent service processing and execution. Edge computing devices are deployed on execution units and collect high-value data for Big Data analytics on the cloud. Through Big Data analytic, cloud computing devices optimize output service rules and deliver the new rules to edge computing devices for better service processing and execution. Both cloud and edge computing are essential to digital transformation of industries.

To demonstrate the importance of cooperation between cloud and edge computing, home appliances are a relevant example. With the popularity of IoT, appliances and furniture can be connected to cloud platforms via the internet. These appliances and furniture can automatically function when the cloud platforms control them. If the network connection fails, they can continue automatic operations. To ensure this, the cloud platforms only control certain functions, while others must be implemented on the device side. This is why edge computing is necessary.

IoT has already emerged and will continue to evolve. That said, many IoT solutions still remain in the Proof of Concept (POC) phase. Though cloud computing has made impressive progress within a short period, edge computing has some catching up to do.

According to statistics from the International Data Corporation (IDC), the world will have over 50 billion terminals and devices connecting to networks by 2020. More than half of data will require analysis, processing, and storage on edge devices. The market of edge computing is huge and imminent.

Intelligent Interconnections of Edge Devices

As discussed, edge computing is the driving force behind the digital transformation of industries. So, what is exactly is edge computing?

Edge computing enables an open platform that integrates with the network, computing, storage, and application core capabilities to provide intelligent services for edge devices near things or data sources. It meets key requirements of digital transformation of industries in terms of agile connection, real-time service, data optimization, application intelligence, as well as security and privacy protection.

Simply, edge computing implements the intelligent interconnection of edge devices. What challenges does edge computing face?

Currently, a high percentage of people connects to networks through multiple terminals. There include smart phones, tablet PCs, and various wearable devices. As the number of connected terminals increases, networks face challenges in connection, O&M, scalability, and reliability. The primary challenges edge computing must address are numerous connections and heterogeneous environment management. In addition to data centers, industrial sites have traditionally utilized various heterogeneous connection buses and Ethernet of different standards. As such, edge computing must ensure devices are compatible and capable of providing real-time and reliable network connections.

Cloud computing and IoT must ensure users are able to access their service systems anytime, anywhere. As such, real-time continuous service interaction is very important, particularly in industrial scenarios. This is because system detection, control, and execution are required. Certain scenarios require delays below 10 ms. Edge computing must meet service requirements of real-time performance.

Construction of the smart society has begun. Service process optimization, O&M automation, and service innovation drive services to be more intelligent. Services are processed more efficiently, with lower costs but a higher level of automation. It is core to edge computing that edge devices become more intelligent. A common application of edge computing is predictive maintenance, a representation of service patterns and business models.

Edge computing must implement service data aggregation and interoperability—the foundation of service intelligence. However, ensuring data aggregation and interoperability across vendors and applications is quite challenging in such heterogeneous environments.

Security risks are omnipresent in both the cloud and edge computing fields. All data centers and terminals require end-to-end protection. Network edge security involves devices, networks, data, and applications. Additionally, requirements include key data security and consistency as well as the protection of privacy.

Edge computing will not be implemented overnight, but will instead require planned development. Its design should be top down, but requires further improvement in connections, intelligence, and autonomy.

Edge Computing Drives and Is Driven by Intelligence

With the continued expansion of digital transformation, edge computing becomes increasingly prominent, and is driven by and drives technology and business. Vendors and users face a major challenge with regards to sustaining and leading the development and application of the edge computing industry.

Edge computing is applicable in many scenarios and adds a great deal of industry value. For instance, it supports the innovation of business models, helping industry users transform from product-oriented to service-oriented. Additionally, it enables customized and intelligent products and services. Edge computing has typically been used in three scenarios thus far: predictive maintenance, energy efficiency management, and smart manufacturing.

In recent events, elevator related injuries have been prominent. Though sometimes due to human error, technology faults were more common as a cause. Currently, over 15 million elevators are in use worldwide, and as such elevator maintenance and after-sale services are a massive industry. To achieve digital transformation and business model innovation, the elevator industry relies primarily on edge computing to improve predictive maintenance and O&M efficiency while reducing O&M costs.

Specifically, built-in sensors can inform relevant personnel of the elevator's running status in real time. Local edge computing with a gateway can provide data analysis capability that predicts device faults instantly. Predictive maintenance can reduce maintenance costs by between 7% and 60%, reduce service interruptions, and increase the lifespan of 70% of devices by 100% to 1000%.

According to reports by Lux Research — a research and advisory company — predicted value of the global Industrial Internet of Things (IIoT) market is US$ 151 billion in 2020. Predictive maintenance applies to not only the elevator industry but also special-use vehicles, computer numerical control (CNC) machines, secondary water supply equipment, and the energy industry.

Edge computing is also widely applicable in the energy efficiency management of public infrastructure. For example, 80% of the world's streetlight vendors are preparing to implement intelligent streetlights. This will reduce energy usage and emissions. Huawei Connected City Lighting Solution can reduce maintenance costs by up to 80%, and ensure multi-level reliability. Normal operations and management can be guaranteed even in cases of cloud control anomalies, ensuring predictive maintenance. Edge computing can effectively reduce building energy consumption.

According to data from the customer in the Australia Melbourne project, the Building Energy-efficiency Management System (BEMS) Solution reduces total energy consumption by approximately 60%. Automatic gathering of energy information eliminates the cost of manual collection while reducing maintenance cost.

Edge computing is also capable of accelerating the implementation of smart manufacturing. Made in China 2025 aims to implement all-round intelligent transformation of the primary sectors of China's manufacturing industry. This change will decrease the operating expenses (OPEX) of trial projects, shorten the production cycles, and lower defective product rates — all by up to 50%. To prepare for the implementation of smart manufacturing, flexible interactions between ICT and OT systems must be improved.

Germany has set an example in implementation of Industry 4.0. For example, with a public cloud on which an IoT platform is deployed, Deutsche Telekom has helped Germany's machinery manufacturing company Holmer implement predictive maintenance on its agricultural machinery.

To date, edge computing has been successfully applied in multiple sectors including public infrastructure, smart energy, smart manufacturing, Internet of Vehicles (IoV), and smart home. This has created solid foundation for the in-depth development of edge computing.

Edge Computing Facilitates a Large Ecosystem

In the cloud computing era, vendors must cooperate to succeed. An ecosystem that facilitates the unification of upstream and downstream partners is required to promote fast and sustainable industry development.

An appropriate ecosystem is also required for edge computing. As edge computing is still emerging, understanding of cloud computing, matching products, and solutions is lacking. Top-level design and operation models of edge computing architecture require further specification, and edge computing industry is still taking shape. Further time and effort are required to integrate administration, production, academics, research, and application resources. Resource integration can effectively accelerate the application of edge computing in various industries while promoting digital innovation.

Research institutions and vendors must actively analyze requirements for edge computing application and summarize key industry requirements to develop conditions suitable for edge computing.

Based on user requirements, the technical reference architecture of edge computing is determined. Software and hardware stakeholders can provide overall solutions based on positioning and division of labor.

Vendors and beta users can accelerate testing and verification of the technical framework, products, and solutions of edge computing. This will accelerate the implementation of edge computing products.

Industry promotion, showcase sites, and experience sharing can improve edge computing awareness, which will in turn accelerate development.

Additionally, stakeholders in the industry should cooperate in research, the formulation of unified standards and regulations, and increasing national and international communications. Openness and collaboration will accelerate the viability of edge computing, adding value to societies and enterprises.

As an emerging industry, the edge computing industry has a broad application prospect. It covers multiple fields including OT, Information Technology (IT), and Communication Technology (CT), and involves many industry chain roles such as network connection, data aggregation, chip, sensing, and industry application. It is reported that Huawei Technologies Co., Ltd., Shenyang Institute of Automation of Chinese Academy of Sciences, China Academy of Information and Communications Technology (CAICT), Intel Corporation, ARM Holdings, and iSoftStone formed the Edge Computing Consortium (ECC) on November 30, 2016. The ECC will be responsible for promoting in-depth industry collaboration and accelerating the digital innovation and industrial application of edge computing.

Edge computing is a key factor for digital transformation. Edge computing alongside cloud computing will significantly affect IT application in various industries.