With the development of the Internet of Things, the amount of data generated by industrial manufacturing equipment will increase dramatically. If all this data is processed in the cloud, it will require endless spectrum resources, transmission bandwidth, and data processing capabilities, inevitably overwhelming the cloud. This is where edge computing comes in to alleviate the pressure on cloud computing.
“90% of the data we collected was garbage,” lamented the owner of a factory in Kunshan, Jiangsu. “Last year, we collected data almost every moment of every day of the year, but we didn’t know how to use the collected data. Compared with the various costs invested in data collection, I don’t think it was worth it.”
A year of data collection experience caused this business owner to lose his initial enthusiasm for the Industrial Internet, and he even began to question: Do we really need a lot of industrial data right now?
"Why should I bother collecting data and developing an industrial internet when a problem that can be solved simply by adding a few more workers? And it might not even be effective!"
Indeed, no matter how impressive concepts like the Industrial Internet of Things (IIoT), big data-driven approaches, and digital twins may sound, in actual industrial production, if they cannot solve a company's core problems—increasing profits and reducing costs—they are merely theoretical. While data itself is important, service applications that directly solve problems are far more valuable to businesses. Currently, besides how to collect data, the key question facing most companies is: what data is worth collecting? Simply put, how to use data to generate value!
We know that industrial data acquisition and transmission are basically carried out in a "device-pipe-cloud" model. In the field of application, the "device" is responsible for collecting data and executing instructions, the "pipe" connects the data transmission path, and the "cloud" is responsible for all data analysis and control logic functions. Whether the entire process can be successfully implemented is crucial to the capabilities of data acquisition, analysis, and application.
However, with the development of the Internet of Things (IoT), the amount of data generated by industrial manufacturing equipment will increase dramatically. If all this data is processed in the cloud, it will require endless spectrum resources, transmission bandwidth, and data processing capabilities, inevitably overwhelming the cloud. This is where edge computing comes in to alleviate the pressure on cloud computing. For example, when a company is small, the board of directors can manage every detail of the company, but as the company grows to a certain size, it needs to grant frontline employees the necessary autonomy.
Therefore, edge computing gateways, which collect, process, and transmit data at the edge of the industrial site, bear the important responsibility of connecting the "Ren and Du" meridians of industrial data transmission. Only by integrating with the cloud platform—edge-cloud integration—and finally using big data analysis to empower production can the true value of industrial data be realized.
This raises two key issues that we must confront:
I. With a large amount of industrial data being transferred to lower levels, how can the validity of the data be guaranteed?
2. What value can edge-cloud integration bring to the Industrial Internet of Things?
"The layman sees the spectacle, the expert sees the details." Inhand Networks, which has been working in the field of Industrial Internet of Things for 17 years, has a lot to say about these two questions.
Boosting edge computing to address the pain points of data sinking.
Gartner's "Top 10 Strategic Technology Trends for 2018: From Cloud to Edge" report states that by 2022, as digital businesses continue to grow, 75% of enterprise-generated data will be created and processed outside of traditional centralized data centers or the cloud.
With the development of the Industrial Internet of Things (IIoT), there will inevitably be more local control and field data. Faced with this increasing amount of field data, how can we process it to ensure its effectiveness while reducing the pressure on cloud computing?
In the industrial world, even the smallest improvement can bring significant advantages; conversely, even the smallest failure can lead to substantial losses. Much data in industrial settings has a short "shelf life"—once processing is delayed, it quickly deteriorates, its value plummeting. Data processing in industrial settings can be described as "walking a tightrope." This is where "edge computing" plays an irreplaceable role.
If we compare the brain to the cloud, then edge computing is like the nerve endings, processing simple stimuli and feeding back the processed feature information to the cloud brain.
While the core issue currently pursued by industrial enterprises is how to empower production and generate value through data, we cannot ignore the common problem that has plagued industrial enterprises for many years in this process: how to collect data—a crucial preliminary step in data processing. For any industrial enterprise, the first step in mining the gold mine of data is data collection. Big data analysis without data collection is a castle in the air, and an industrial cloud platform without data is like a tree without roots.
In different industrial production processes, due to the large number of automation product brands, the diversity of industrial interfaces, and the lack of unified industrial protocols, data acquisition, which seems simple, is not so easy.
In addition to data collection, in terms of data processing and application, industrial site data faces the problem of a very short "shelf life" and a large amount of "garbage" data that does not need to be transmitted to the cloud.
While edge computing is developing rapidly from an industry perspective, it is still in its early stages of practical application. The true value of industrial data can only be realized through the integration of edge computing and cloud computing.
In fact, the industry has already recognized the importance of edge-cloud collaboration and has been actively exploring it. For example, Huawei, at its HC2018 conference, released the Intelligent Edge Platform (IEF), which explicitly proposed the concept of integrated edge and cloud collaboration services; Siemens released the Industrial Edge concept in 2018, the general idea of which is to achieve collaboration between edge computing and cloud computing by deploying Industrial Edge Management in the cloud; and InHand Networks, at this year's Hannover Messe, showcased "InHand Device Networks Cloud + Edge Computing Gateway" based on an edge computing gateway, realizing edge-cloud collaboration.
Since edge-cloud collaboration is so important for industrial data, how do we understand it? The key to edge-cloud collaboration in processing data lies in data fusion.
In industrial scenarios, edge computing enables real-time analytics algorithms to run directly, while the collaboration between edge and cloud facilitates continuous model growth and optimization, thereby enhancing the platform's real-time analytics capabilities. Of course, the specific capabilities and focus of edge-cloud collaboration will differ across application scenarios, as each edge computing business model has varying requirements for collaboration with cloud computing.
For example, in flexible manufacturing, the application of modern industrial robots is becoming increasingly widespread. The stability and reliability of robots and robotic arms on the production line are of great significance to ensuring the economic benefits of enterprise production. The large-scale deployment of industrial robots, their complex structure, and high maintenance costs place extremely high demands on the maintenance capabilities of technical personnel in production enterprises. This is mainly reflected in the need to detect abnormalities in robot components and control devices before robot failures occur, and to alert users to perform targeted maintenance and repairs before downtime occurs, thereby reducing downtime to zero and achieving continuous production.
The key here is to achieve continuous and effective production through preventative maintenance via edge-cloud collaboration.
In the cloud, the equipment cloud can collect real-time production data from industrial sites for centralized storage, analysis, processing, and prediction. From network management and on-site detection to sensing and response, it can greatly improve operation and maintenance efficiency.
In conclusion, comparing data to oil today is no exaggeration. Just as oil needs to be collected, transported, processed, and refined before it can be used, so too does industrial data. Edge computing offers more powerful insights and analytical capabilities for collected data. The application of edge computing and the deployment of edge computing gateways will make the benefits generated by the data clearly visible, dispelling factory owners' doubts about industrial data, and enabling the Industrial Internet to truly be implemented at the "front line."
Edge-cloud collaboration has brought immeasurable value to ICT vendors, OT vendors, OTT vendors, and telecom operators. Through in-depth data mining, it promotes business innovation and business model innovation, and accelerates digital transformation.
In the era of intelligent manufacturing, all aspects of production need to be interconnected and able to interact in real time. For example, production data and equipment data from production, warehousing, and logistics need to be monitored and tracked in real time, and then intelligent predictions can be made through big data processing, including advance inventory preparation and safety precautions. Inhand Networks' industrial IoT deployment follows the footsteps of Industry 4.0, based on "edge computing gateway + device cloud + big data analysis." It adopts edge-cloud collaboration to open up channels for data collection, transmission, and processing, and performs big data analysis to fully leverage the value of data, ultimately empowering the industrial IoT in all aspects.
