The Industrial Internet can be used to reconstruct the ecological relationships between enterprises, as well as to facilitate the transformation and upgrading within large enterprises. Today, I would like to share some thoughts on large enterprises: What are the differences between the Industrial Internet and the past automation and information systems? I have spent five years pondering this question and have finally come to an understanding.
Twenty years ago, Director Wu of the Baosteel Continuous Casting Workshop had an idea: to extract data from the continuous casting equipment and transmit it over the network. This way, the equipment status could be monitored from the office. In hindsight, this is a typical Industrial Internet mindset. However, I was not enthusiastic about this proposal and did not participate. The reason is simple: while this approach would make it easier for inspection workers, the engineering effort required is still considerable. I felt that the return on investment was not worth it. According to the company’s evaluation system, engaging in such work would basically be a futile endeavor.
In fact, the confusion enterprises have regarding the Industrial Internet centers around the word “value”: What exactly can the Industrial Internet do?
More than 20 years of innovative work have accustomed me to consider value from the perspective of “opportunity”: What things were previously impossible or unnecessary to do, but are now feasible and necessary? Opportunities often lie hidden here. Therefore, my reflections often begin with a review of history.
It is well known that the computer systems within large enterprises are often quite comprehensive. However, these systems are usually hierarchical: the lower-level computing systems are generally referred to as automation systems (such as BA, PCS) while the upper-level systems are called information systems (such as ERP). The lower-level computers typically handle real-time issues with a very high degree of automation; whereas the upper-level computers deal with less time-sensitive issues that require a higher proportion of manual management. However, the impact of automation systems is quite limited, affecting only a valve or a piece of equipment; while information systems can manage a workshop or an entire factory.
In principle, there should be information links between the various levels of computer systems. The upper-level computers are responsible for generating goals and requirements, which are then transmitted downwards and decomposed for execution by the lower-level computers. The execution status from the lower-level computers is also uploaded to the upper-level computers to allow managers to grasp the production situation. However, this uploading is selective, generally only transmitting production results or special issues. Data uploads rarely trigger automatic feedback and adjustments from the upper-level computers.
So, the question arises: Where are the opportunities for the Industrial Internet?
In my view, this opportunity lies in the contradiction mentioned earlier: the lower-level systems have strong real-time capabilities but a small management scope, while the upper-level systems have poor real-time capabilities but a large management scope.
Theoretically, achieving automation is ideal. We know that one of the roles of automated control systems is to replace some human tasks. However, the significance goes beyond that. For instance, the response time of automated control systems can easily reach the millisecond level, whereas human response times are significantly slower. Therefore, the control effectiveness of automated systems is often superior to that of humans, and their value in improving quality and reducing costs is even more important than merely reducing labor costs.
That being said, why can’t the decision-making of information systems be automated? One important reason is that the issues faced by information systems are often much more complex than those dealt with by automated control systems. For example, the automated control system functions under the premise that the equipment is fault-free; however, when an issue arises with a piece of equipment on the production line, managers need to determine how to respond. Some problems can only be flexibly handled by humans. Management systems are essentially designed to assist humans in dealing with problems flexibly.
Humans have their strengths as well as their limitations.
The collection of equipment information can be at the millisecond level, and an entire factory could have thousands of devices. This equipment information not only includes the status of the equipment but also the operational and management information from humans. If there is enough information, various issues related to quality, efficiency, and cost can be identified from the equipment data. However, humans are not omniscient. If all this data is sent to humans for processing, their real-time processing capabilities would be overwhelmed.
What should we do when human brains are insufficient?
In GE’s white paper on the Industrial Internet, three key factors are proposed: intelligent machines, advanced analytics (algorithms), and (online) human workers. Among these, “advanced algorithms” essentially allow computers to act as human assistants; they help people monitor and analyze this data, escalating issues to humans only when necessary. In this way, humans can leverage algorithms as an “external brain” to better manage the factory. It is through the collaboration of humans and algorithms that problems with vastly differing time cycles can be addressed. In fact, the subtitle of GE’s Industrial Internet white paper is: Pushing the Boundaries of Minds and Machines, which can be understood as “restructuring the interface between humans and machines,” reflecting the collaboration between humans and machines.
Specifically, the machine as a “secretary” possesses many capabilities that surpass human cognition: its memory capacity is exceptionally strong. If the historical data of thousands of devices is collected, it constitutes a significant body of knowledge. This is something that the human brain cannot handle. Five years ago, when I invited Mr. Salvo to speak at Baosteel about the Industrial Internet, he mentioned that the Industrial Internet could assist people in designing products in parallel, leading to intelligent product design and development processes. It was only recently that I truly grasped this issue.
When humans collaborate with “advanced algorithms,” it is akin to a pedestrian boarding an airplane. The ability to manage enterprises is greatly enhanced. I have repeatedly emphasized: “Management defines the boundaries of technology.” Once management capabilities are enhanced, the competitiveness of enterprises naturally rises: costs can be significantly reduced, quality and efficiency can be greatly improved, inefficiencies can be eliminated, and corporate culture can become healthier. The value is immeasurable. Thus, the confusion surrounding the “value issue” of the Industrial Internet is resolved.
Let us return to Director Wu’s idea from 20 years ago. At that time, not only was I uninterested in his idea, but the company was also disinterested. If we could go back 20 years, is there a way to spark the company’s interest? Yes! This method is to help the company articulate its strategic vision. The value of this endeavor should not be viewed solely from a singular perspective but rather as a means to become the most competitive steel enterprise in the world and to cultivate Baosteel’s future competitiveness in the IT sector. Framing it this way may elevate the importance of this task among everyone. Of course, the chosen entry point does not necessarily have to be continuous casting.
I often say: Innovation often originates from personalized needs. Great innovations often emerge from seemingly trivial beginnings. It now seems that without viewing problems from a strategic perspective, the winds of change cannot be stirred.
In summary, the functions of the Industrial Internet differ from traditional automation and information systems in enterprises. This distinction is where the opportunities of the Industrial Internet lie. In the short term, the Industrial Internet systems should perhaps operate in conjunction with existing enterprise systems, cooperating and collaborating. This approach helps mitigate the risks associated with system upgrades. When conditions are ripe, it is also anticipated that a unified system will be used to integrate the functionalities of various systems.