5G cannot enhance industry?

There are already more than 1,100 “5G+Industrial Internet” integrated innovation projects.

This is a set of data officially given at the first 5G+ Industrial Internet Conference in China a few months ago.

If extended to April 2021, with the support of multiple policies and the promotion of 5G+ by telecom operators and ICT equipment manufacturers, this number may have more than doubled.

But under the halo of 5G+, some practitioners in traditional industries are a little impatient.

"Those examples of public speeches and presentations are more or less glamorous. But in fact, traditional steel companies like us have many historical problems left over from decades. We have not formed systematic interconnection and interoperability. In addition, various equipment is relatively bulky, and the information security line is relatively weak. The so-called digital transformation is still an attempt at some local links. Some industrial control systems still use the systems of foreign companies. Incorporating 5G and building the industrial Internet will still take a long time to connect and explore."

When a technician from a steel company talked to Zijinshan Technology about industrial upgrading and transformation, he said that most of the current 5G+ cases are "false fire" and "have not reached the level of prosperity created by various propaganda."

[[396267]]

01 Standards and frequencies: invisible fences?

Industry is regarded as the industry field that best reflects the major characteristics of 5G, such as high speed, low latency, high reliability, and large connections. It has always been the "frontline" for the three major operators and equipment manufacturers to test the waters of 5G+.

But behind the publicity cases, when factories and enterprises began to spend a lot of energy and money to "comprehensively transform" various links such as procurement, production, quality inspection, and logistics, 5G, which was restricted by standards and spectrum policies, began to show its inability to keep up with the times in terms of technical capabilities and service models.

In fact, the 5G SA independent networking standard frozen in June 2018 has significantly increased the speed and capacity of 5G networks, opening the door for it to enter the industrial market. The 3GPP R16 standard was fully frozen in July 2020. Because it unlocks newer capabilities for 5G+industry, it is regarded as a stepping stone for 5G in the industrial field.

How many technical capabilities can truly “transform” industrial production are included in the released R16 standard?

Professor Wang Jianquan, an industrial Internet expert and a member of the National Thousand Talents Program of the University of Science and Technology Beijing, said when discussing this issue with Purple Mountain Technology that the 3GPP R16 standard does enhance and expand 5G-related functions, including a framework definition of the architecture and functions of 5G bridging TSN, enhancements to the high-reliability and low-latency uRLLC function, and planning for the deployment of non-public network NPN, including NR-U, NR positioning, dual connectivity, carrier aggregation enhancements, and UE energy saving.

However, if 5G wants to better adapt to complex industrial scenarios in terms of latency, capacity, connection, positioning accuracy, etc., it still has to rely on the R17 standard that has not yet been promulgated.

On the one hand, the more powerful R17 standard has not yet been released, and on the other hand, the actual progress of 5G is still lagging behind the released standards. According to the roadmap of terminal chip manufacturers that have been made public, the more comprehensive and mature uRLLC terminal chips are expected to be launched from the end of 2021 to the middle of 2022, and the modules based on this aspect will be delayed by several months.

Professor Wang Jianquan said frankly, "For scenarios with high requirements for latency, jitter, and stability, especially the core control links of industrial manufacturing, the current 5G cannot meet the requirements. This is also the technical reason why 5G has not yet entered the industry and has not been truly linked to the core processes of production and manufacturing."

Professor Wang Jianquan and other interviewees also mentioned another major reason why 5G+ industry is difficult to promote besides technical factors, namely the business model, which is closely related to my country's existing 5G frequency policy.

As an industry insider said in private exchanges with peers during the MWC Shanghai exhibition in February, China has the most 5G+ cases in various industries internationally, but contrary to the fact that "37 countries and regions in the world have completed the spectrum allocation and deployment of dedicated networks based on 5G/LTE", China has not yet allocated dedicated spectrum for 5G private networks.

"It's like a big playground. There's no dedicated track for professional athletes, so they have to run with ordinary people."

Germany and Japan both allocated specific spectrum for 5G private networks in 2019. Germany's planned 3.7GHz-3.8GHz and 24.25GHz-27.5GHz frequency bands have been allocated in batches in the past two years for the construction of 5G enterprise private networks in industrial plants, automobile companies, agricultural and forestry companies, etc. According to statistics, by the end of September 2020, dozens of companies including Bosch and BMW had received 74 regional 5G spectrum licenses to build 5G independent private networks.

The 2.575MHz-2.595MHz and 28.2GHz-28.3GHz frequency bands issued by Japan are specially marked as "locally dedicated", that is, local 5G frequencies can only be used in local private networks, and national operators cannot apply for local 5G frequencies. Fujitsu, Toyota, and the Tokyo Metropolitan Government have all obtained local 5G frequency licenses and have begun to build regional 5G private networks.

The United States and France also use different spectrum sharing access regulatory models or independent frequency bands for industrial enterprises to build private networks.

In the view of experts from the Research Institute of China Telecom Corporation Limited, the frequency allocation of industrial private networks in these countries has more or less provided inspiration for the domestic promotion of 5G private network spectrum: first, compared with the national 5G public network spectrum of telecom operators, most private network spectrum has limited usage areas; second, private network spectrum can be allocated to enterprises through secondary authorization by regulatory authorities.

Due to the limitation that there is no 5G private network spectrum in China, cooperation with telecom operators has become the main way for industrial enterprises to build 5G networks. Telecom operators have also introduced different cooperation modes such as network slicing and co-construction.

Network slicing means that operators are responsible for network operation support and do not need enterprises to deploy localized network hardware infrastructure. It is suitable for enterprises with low requirements for data security to deploy 5G private networks.

Co-construction means sharing or leasing the operator's public core network, building an access network within the enterprise park, and sinking the local core network, but this also poses risks for companies with extremely high security requirements.

Regardless of which method is adopted, there are limitations on the 5G public network spectrum for various complex demands in industrial production, as well as high costs in terms of construction, maintenance, and services.

Previously, the China Industrial Internet Industry Alliance investigated the spectrum requirements in two typical industrial scenarios: automobile manufacturing and electronic information manufacturing. The spectrum requirement for the automobile manufacturing industry scenario is between 200M-600M. When using omnidirectional cells, the spectrum requirement is as high as 1800M; the spectrum requirement for electronic information manufacturing factories is between 60M-350M.

Just like in a large playground that is already limited in space, companies still have to find ways to carve out their own track.

Large companies that are "not short of money" prefer to rely on dedicated spectrum companies to build their own.

When an industry insider "Tianquanquan" mentioned this issue, he said that although the cost of building a 5G private network is very high, the way of building it together with operators also means that certain expenses are required in terms of services. Large companies with experience and capabilities like Foxconn are likely to prefer building their own 5G independent private networks based on security, physical isolation and other requirements. Therefore, many car companies, electronic manufacturing companies, etc. with technical capabilities tend to build their own private networks, unless the company has wide-area connection needs (such as power grids, Internet of Vehicles), and the cost of building their own networks will be very high.

02 Operators hovering on the periphery

"In the automation transformation of China's industrial production, it took nearly 20 years to transform a relay into an IGBT, but mobile communications have developed from 2G to 5G in the past 20 years. We often joke that a big reason for people to divorce is that one person has gone too far, while the other has stayed where he is, resulting in the two no longer having a common language. What the industry needs is stability and no change, while the development of the communications industry is constantly changing and iterating rapidly. Now 5G and industry seem to be in this situation."

A Zhihu user named "Lai Zhenbo" described the difficulty of 5G entering the industry.

For decades, the business model of telecom operators has also prevented them from truly and perfectly connecting with the complex needs of industrial enterprises.

"A large number of robots have begun to be introduced in industrial production and quality inspection. 5G networks can replace wired networks to greatly improve the efficiency of information transmission. Robots have sensor data collection, which requires large-capacity wireless transmission. The issued control commands also need to be on a more reliable network and reach the destination with lower latency. Therefore, compared with the card issuance, package marketing, and traffic selling that operators are better at, what industrial enterprises need is an intermediate service provider who understands industrial enterprises very well, from architecture, data platform, AI big data analysis, data connection and sensors to data network integration and then to production systems to achieve digital transformation." The above-mentioned person said that this is also the biggest gap between operators and industrial enterprises at present.

In addition, the “Antenna Circle” also mentioned that many links in the industry have extremely high requirements for uplink data bandwidth, such as automatic control and real-time monitoring by high-definition cameras.

"In fact, telecom operators have also provided some technologies to enhance uplink capabilities, such as 5G super uplink and uplink dedicated frame structure, which can meet the uplink bandwidth needs of industrial enterprises to a certain extent. However, the industrial chain in this area needs to mature, and base stations, terminals, and frequency bands all need to support such technologies."

This also involves the demand for higher frequency bands such as millimeter waves, but currently there is a lack of licensing support and complete industrial chain capabilities.

A person from Midea Group also mentioned the issue of 5G modules. "Not to mention the cost of building 5G networks for enterprises with a scale of millions, tens of millions, or even hundreds of millions, the price of 5G industrial modules is also high. Moreover, there are too many 5G modules at present, and there is no room for more in the motors." Therefore, the person also called on module manufacturers to jointly launch a more reasonable solution with the three major operators.

This reality has forced telecom operators to invest heavily in 5G+ on the one hand, but on the other hand they have never been able to enter the core links.

Moreover, there is another reality facing 5G+ Industry, that is, more than 90% of enterprises in my country's manufacturing industry are small and medium-sized enterprises. Most of these enterprises' existing equipment does not yet have the ability of intelligent digital collection. The financial and cost pressures of upgrading and transformation are very high, and enterprises are unable to bear it.

Even if we can handle 10% of the large enterprises, how can we serve the 90% of small and medium-sized manufacturing enterprises?

Perhaps, the test of whether 5G can be fully integrated into industry is not just a matter of standards, spectrum, technical capabilities, and cooperation with intermediate service providers.