What is the Industrial Internet ? A hundred people will have a hundred different interpretations. Everyone has their own Hamlet, and everyone has their own Industrial Internet. Every organization has given its own definition or connotation of the Industrial Internet.
I. Industrial Internet: "Connectivity-Control-Optimization-Efficiency"
GE defines the Industrial Internet as: "a continuous process that occurs through the integration of global industrial systems with advanced computing, analytics, low-cost sensing capabilities and new levels of connectivity brought about by the Internet."
Japan's "Interconnected Industry" is an industrial transformation and upgrading activity that is highly similar to the Industrial Internet. Its essence is to create new value and generate new technologies, new products and new service models by realizing the interconnection of things, machines and equipment, enterprises, people, data and even production and consumption.
The China Industrial Internet Alliance defines the Industrial Internet as follows: The essence of the Industrial Internet is based on the network interconnection between machines, raw materials, control systems, information systems, products, and people. Through comprehensive and in-depth perception, real-time transmission and exchange, rapid computing and processing, and advanced modeling and analysis of industrial data, it achieves intelligent control, operational optimization, and transformation of production organization methods.
The Industrial Internet Research Institute of the Ministry of Industry and Information Technology believes that the Industrial Internet is a product of the deep integration of new-generation network information technology and manufacturing. It is an important infrastructure for realizing the digital, networked and intelligent development of industries. Through the comprehensive interconnection of people, machines and things, and the comprehensive linking of all factors, the entire industrial chain and the entire value chain, it promotes the formation of a brand-new industrial production, manufacturing and service system, and becomes a key support, important path and new ecosystem for the transformation and upgrading of the industrial economy.
There are many other definitions and explanations, which will not be elaborated here. Although the views of different parties may not be exactly the same, the consensus is that connecting machines and equipment, connecting industrial systems, and connecting various industrial elements (industrial "ends") is one of the basic connotations of the Industrial Internet.
Regarding the Industrial Internet, in my article "How the Industrial Internet Promotes High-Quality Development of Manufacturing," I offered my insights: Today's Industrial Internet is an industrial network based on Industrial Ethernet and continuously integrating Internet technologies. Undoubtedly, the Industrial Internet primarily connects industrial elements, linking machinery and equipment within factories and industrial systems. In that article, I proposed five basic logics for achieving "connectivity-control-optimization-efficiency" through the extensive connection of machinery and equipment with industrial systems:
Because it connects industrial equipment across time and space, it can monitor and manage a wide range of machines and equipment and enterprise supply chains in real time, and even extend the control to smart products in the hands of customers.
Because comprehensive and accurate data collection, processing, and traceability are carried out, accurate and reliable data can be used for product models, business models, and market forecasting, supporting various calculations and decision-making processes for enterprises.
Because data flows automatically and "does not land," it ensures that the right data, at the right time, and in the right version, is delivered to the right people and machines, allowing things to be done right the first time and optimally.
Because the software has formed a closed-loop control, it can enable machines to operate autonomously. When humans leave the system loop, machines can do the same thing without human intervention, or even do it better.
Because it allows for the optimization and allocation of manufacturing resources on a large scale, it enables large-scale enterprises to achieve refined and intensive management, and realize overall optimization at the level of complex systems.
(The above views are based on the author's article "How the Industrial Internet Promotes High-Quality Development of Manufacturing")
Therefore, the starting point for the benefits of the Industrial Internet is connecting industrial equipment and systems, and the endpoint is sending the control commands generated in the cyber system to the controllers of the machines to precisely control their operation. Thus, the primary task of the Industrial Internet is to achieve interconnectivity between various machines and industrial systems! If someone evades the question, beats around the bush, goes on and on, spouting a lot of high-sounding "advanced jargon" and introducing numerous solutions for connecting computers, mobile phones (and the people behind them), and classic manufacturing information software, but never discusses how to connect various machines and industrial systems, then they are definitely avoiding the most important issue and are certainly unable to provide a true Industrial Internet solution.
The history of the Industrial Internet can be traced back to the Internet of Things (IoT) of devices that began in the last century.
II. Five Stages of Development of the Internet of Things (IoT) for Devices
In 1969, Digital Equipment Corporation of the United States developed the world's first programmable logic controller (PDP-14); in 1971, Japan developed the first programmable logic controller (DCS-8), thus ushering in what the Germans defined as the "Third Industrial Revolution".
A Programmable Logic Controller ( PLC ) is a digital control device that evolved from the relays widely used during the Second Industrial Revolution by incorporating computer technology. It has an internal programmable memory that stores instructions for performing logic operations, sequential control, timing, counting, and arithmetic operations. Through analog or digital signal input and output, it controls various types of machinery or production processes.
In their co-authored book, *Ingenuity*, Zhu Duoxian, Chairman of Langguang Innovation Company, and I mentioned that DNC is generally referred to as machine tool networking. In a narrow sense, DNC refers to using a single computer to transmit programs between multiple machine tools. In a broader sense, DNC encompasses program network transmission, program editing and simulation, program management, equipment monitoring and data acquisition, big data analysis and report display, etc. It is a comprehensive system for the networked management of CNC equipment and an indispensable information management system for machining workshops.
MDC (Manufacturing Data Collection) is a system that collects manufacturing data. Since most MDC solutions on the market currently focus on machine tools, it is often referred to as a machine tool monitoring system. MDC uses advanced hardware and software acquisition technologies to collect real-time, automatic, objective, and accurate data from CNC equipment, enabling transparent management of the production process and providing automatic feedback of production data to the MES system. MDC can function as a standalone application system or be used in conjunction with other DNC modules; it can be considered part of a broader DNC system.
Device networking and data acquisition systems are generally called DNC/MDC systems, or device Internet of Things (IoT) systems. Of course, there are differences between device IoT and DNC/MDC. Device IoT has a wider and deeper range of applications than DNC/MDC, and can be considered a new stage of development for DNC/MDC.
Today, DNC/MDC has evolved from simply managing the network of CNC machine tools into a system for managing the network of various digital devices. It connects and collects data from digital devices such as CNC machine tools, robots , testing equipment, and heat treatment equipment, and is a specific application of CPS (Cyberphysics System) in the workshop.
Figure 1 illustrates the five stages of the development of the Internet of Things (IoT) for devices.
Figure 1. Five stages of development of the Internet of Things (IoT) for devices.
(The above text and images are from "Ingenuity: From Digital Workshops to Smart Manufacturing")
This brief review shows that the Internet of Things (IoT) for devices began its development in 1969, coinciding with the development of ARPANET, the progenitor of the Internet. IoT and the Internet can be considered to belong to the same family of Internet applications, forming a large family of Internet applications. It is precisely because IoT and the Internet developed in parallel over time and mutually learned from each other technologically that Industrial Ethernet, based on TCP/IP (Transmission Control Protocol/Internet Protocol), was formed.
In terms of applications, Industrial Ethernet and the Internet of Things (IoT) for devices form the Industrial Internet, while the Internet forms the Consumer/Social Internet. Regarding the differences and connections between the two, Yin Jinguo, General Manager of E-Tech, uses this analogy: "From a technological development perspective, I believe the Industrial Internet and the Consumer Internet are not about who came first or last, or who controls whom. Rather, they are two different flowers growing from the same soil. Both flowers are beautiful, each with its own unique fragrance, but they are two different species. Only in the end, when their branches are lush and verdant, do they intertwine, and the space where they intertwine is the user and the market."
Since Industrial Ethernet and IoT devices are used to connect machines and industrial systems, how to connect these machines becomes the most fundamental and crucial element for the implementation of the Industrial Internet. Connecting machines presents numerous challenges and pain points—insurmountable "Mount Everests" that must be climbed.
III. Challenges in Networking Machines and Industrial Systems
Regardless of the manufacturing model/paradigm or the level of advancement of machinery and equipment, equipment connectivity and data collection remain the most practical and frequent needs in production, and prerequisites for the "networking" of the Industrial Internet. If machinery and equipment cannot be connected, data cannot be collected correctly, and effective data analysis is impossible, then those who talk about the Industrial Internet but have never actually worked on connecting machinery and equipment should basically just go away and rest!
Networking machinery and industrial systems involves numerous industrial elements and covers a wide range of sectors. Typically, industrial systems such as factories, municipal facilities, oil and gas plants, power plants, mines, logistics companies, automotive companies, large equipment manufacturers, and buildings are all involved and fall within the scope of data collection.
Through visits to enterprises, the author found that effectively connecting various industrial elements presents a variety of problems and several challenges that need to be overcome:
Challenge 1: The equipment itself does not "produce" data. Some early physical machines, or those initially designed not to require digital modules, such as large industrial boilers and port machinery, as well as small sewing machines and electric irons, lack sensors and computing cores (chips). Adding sensors, chips, or other digital modules will present many unforeseen challenges.
Challenge 2: Closed Digital Modules. Some industrial equipment has built-in digital modules, but these modules are not accessible. The data is stored internally and hidden, with no physical interface for reading it. Forced disassembly could damage the equipment.
Challenge 3: Tampering with digital modules. Some industrial equipment has its own digital modules and interfaces that can read internal data, but these modules have been tampered with by the original equipment manufacturer (e.g., multiple devices are set to the same URL), effectively blocking the possibility of multiple devices connecting to the network at the same time, making it impossible to effectively read the device data.
Challenge 4: Encrypted Digital Modules. Some industrial equipment has its own digital modules and interfaces to read internal data, but the data is encrypted. Without a decryption module, the data format cannot be recognized. For example, a company purchased 600 advanced flat knitting machines, each costing over 900,000 RMB. However, this huge investment does not guarantee normal operation; the company also has to pay 6 million RMB annually (10,000 RMB per machine) for "communication service fees" and data decryption module usage fees!
Challenge 5: Different Buses and Protocols Even without the aforementioned challenges, different buses and protocols remain a major obstacle to device networking. As is well known, since the reform and opening up, due to historical reasons, China's industrial equipment has been a diverse mix of international brands, with a plethora of existing equipment, creating a fundamental national condition that is drastically different from that of Europe, America, and Japan (whose equipment is concentrated on one or two buses and a few common protocols).
Internationally, four major camps of industrial Ethernet technology have emerged. The ones primarily used in discrete manufacturing control systems are: Modbus-IDA Industrial Ethernet; Ethernet/IP Industrial Ethernet; and PROFInet Industrial Ethernet. The one primarily used in process manufacturing control systems is: Foundation Fieldbus HSE Industrial Ethernet.
However, different companies in different countries, driven by their own interests, are unwilling to fully comply with any particular industrial Ethernet protocol. Instead, they modify existing protocols to benefit their own products, forming their own fieldbus technologies. To date, no one can say with absolute certainty how many types of fieldbuses exist in the industrial control field. One estimate is around forty types, while another says over seventy. However, according to Zheng Bingquan, General Manager of Beijing Yacon Technology Co., Ltd., based on his over twenty years of experience in device networking, he has seen all sorts of internationally-branded equipment in China, with nearly two hundred variations of fieldbuses and over 5,000 different device driver protocols encountered.
Although major manufacturers in various countries have been calling for a unified fieldbus and device driver protocol—for example, Germany's industrial equipment integration proposed the OPCUA protocol, while North America and Japan proposed the MTConnect protocol—the path to unification has already begun. However, due to the intertwined interests of various parties, this path to "unification" is destined to be very long. Therefore, given China's national conditions, without a solid grasp of thousands of device driver protocols, one truly cannot handle the delicate task of networking machinery and equipment.
IV. Pain Points in Data Collection and Analysis After Network Connection
Connecting devices to the network is just the first few steps in a long journey. The next step involves collecting massive amounts of data to provide raw data support for digital business systems such as factory monitoring centers (SCADA/HMI), energy management systems (EMS), manufacturing execution systems (MES), and enterprise resource planning systems (ERP).
The collection, management, forwarding, and sharing of industrial data has always been a challenge for the manufacturing industry. In practice, many industrial sites face problems such as a wide variety of data interfaces, difficulty in standardizing protocols, and a lack of universal architectural design when dealing with machine tools, special equipment, robots, subsystems, special modules, controllers, instruments, circuit boards, and power and building systems. Therefore, three major pain points are typically encountered when networking and collecting data:
Pain Point 1: Data Acquisition Issues. Because production schedules in industrial settings cannot be disrupted, slow research progress, high development difficulty, and long verification and maintenance cycles are common problems encountered on-site. These issues severely impact the development and implementation of data acquisition projects, especially given the numerous hidden costs associated with verification and maintenance, and the potential for unexpected problems. Figure 2 illustrates the issues involved in networking equipment in industrial settings.
Figure 2 Data acquisition issues in industrial field equipment networking
Pain Point 2: Hardware and Software Compatibility Issues. Traditional DTUs (Data Transfer Units) can only handle serial port data forwarding and have a low acquisition frequency. They are mainly used in low-frequency data acquisition scenarios such as water supply and heating, where a 1-2 second delay is acceptable. Industrial gateways can handle data forwarding via serial ports and Ethernet links, but they support limited protocols and are difficult to form a complete solution. Traditional SCADA software supports many drivers, can forward data externally, and has a complete solution, but the deployment cost of the host computer is relatively high. Therefore, there are still many hardware and software compatibility issues in the networking of industrial field devices, as shown in Figure 3.
Figure 3. Hardware and software compatibility issues in industrial field device networking.
Pain Point 3: How to effectively utilize existing data? How to achieve the sharing, distribution, analysis, and collaboration of IoT data from devices? How to unlock the value of massive amounts of real-time/historical data? How to move beyond remote monitoring and parameter configuration to coordinating resource allocation, improving manufacturing processes, optimizing production scheduling, and perfecting quality traceability systems on a larger scale, thereby creating value for the enterprise?
The above three pain points severely restrict the speed and quality of data collection and processing.
V. Making the Industrial Internet Easily "Networked"
To overcome the above pain points and difficulties, and to better connect various machines and equipment, there should obviously be software called an "industrial data platform" to achieve the following goals: to benefit enterprises, improve their profitability, and make the industrial internet a "network":
(1) Deployment flexibility: The platform is independent and scalable, and can be flexibly distributed across multiple sites;
(2) Data integrity: Multiple redundancy mechanisms and comprehensive data storage;
(3) Project scalability: The scale of the project can be expanded or reduced at will through upgrades and renovations.
(4) Interface uniformity: rich and diverse interfaces, and standardized data sources.
Such industrial data platform software possesses both depth, grounding itself in on-site equipment and production processes; breadth, ensuring compatibility with various subsystems and providing holistic solutions; and height, enabling insights and predictions about equipment behavior based on collected data. Most importantly, it eliminates obstacles to equipment connectivity for enterprises, providing an easy-to-use solution.
During my recent research, I discovered that such an industrial data platform has been developed and is being used in the practical application of industrial internet-enabled machinery and equipment. This industrial data acquisition platform has the following characteristics:
Extensive equipment support
Supports driver protocols for over 1500 manufacturers and over 5000 types of devices;
Supports PLCs, DCSs , intelligent modules, intelligent meters, boards, frequency converters and other devices;
Supports multiple connection methods including COM, TCP, UDP, GPRS, programming port, and USB;
Supports standard protocols such as OPC, Modbus, Bacnet, Lonworks, IEC101, IEC104, and DNP;
Supports the collection of data from third-party systems;
Supports customized development of non-standard equipment data acquisition.
Multiple redundancy mechanisms for rapid switching: dual IOServer, dual devices, dual network redundancy;
Cross-regional and cross-network data collection and transmission;
It can support the collection of 50,000 variables per second and can run stably for a long time;
Multiple projects can run simultaneously on the same computer;
It can be used as an OPCUA client to collect third-party data.
In my opinion, the support for over 1,500 manufacturers and over 5,000 device driver protocols (basically all existing "international" devices in China can be connected), support for all standard and non-standard protocols, and long-term stable operation, all of these perfectly meet the networking needs of the industrial sector and clear away the technical obstacles for the implementation of the Industrial Internet.
VI. Summary
The "first principle" of the Industrial Internet is to widely connect various machines, equipment and industrial systems to achieve the basic logic of "connection-control-optimization-efficiency": data collection is achieved through connection, equipment monitoring and behavior insight are achieved through data collection, and thus manufacturing resources are optimized and allocated in a well-founded and refined manner.
Connecting machinery and equipment requires a deep understanding of industrial elements and a strong technical foundation, necessitating the overcoming of various connectivity challenges. Familiarity with numerous fieldbuses and mastery of thousands of device driver protocols are crucial for connecting machinery and equipment. Without a solid grasp of thousands of device driver protocols, one simply cannot handle the complex task of networking machinery and equipment.
After connecting machines and equipment, establishing a stable data acquisition process and effective data analysis and insights are key points for implementing the Industrial Internet.
No matter how good the concept of the Industrial Internet is, it still needs to be put into practice by connecting various industrial devices.
No matter how prominent the slogan of the Industrial Internet is, it still needs to create value through large-scale optimization of industrial resources.
The Industrial Internet is designed to connect industrial elements. Any "Industrial Internet" that cannot effectively connect machinery and industrial systems is a "fake Industrial Internet" or a "pseudo-Industrial Internet".
Disclaimer: This article is a reprint. If it involves copyright issues, please contact us promptly for deletion (QQ: 2737591964). We apologize for any inconvenience.
