The Past and Present of Automation Technology

The Past and Present of Automation Technology

At present, one of the key forces driving the vigorous development of the industrial internet is the continued advancement of automation technology. It can be said that automation technology accompanied the rise and development of the first and second industrial revolutions, and played a critically important role throughout.

More than 150 years ago, the earliest process control technologies were closely tied to machine control. Engineers at the time used local hydraulic transmission and pneumatic control methods, enabling local equipment to be controlled through closed-loop circuits. It was from that point that modern control theory gradually began to take shape. Later, with the large-scale adoption of electrical equipment during the second industrial revolution, automation technology made major strides in the field of analog signal control, giving rise to standard current analog signal interfaces such as 0–10 mA and 4–20 mA, alongside the rapid development of process control theory and automatic control theory.

However, control technology truly took off in the 1970s, when large-scale integrated circuit technology began to see widespread use, especially after computer technology started entering the fields of aviation, aerospace, and the military. Only then did automation control technology enter a period of rapid growth. Many emerging automation companies rose quickly during that time, including industry giants that still wield enormous influence today, such as Honeywell, Yokogawa, Schneider Electric, GE, and Siemens, as well as firms like Bailey and Rosemount, which have since faded from view.

The control technologies that emerged in the late 1970s resembled the computer systems of that era: centralized control was the dominant model, with all field signals collected in a central location.

But the drawbacks of centralized control became increasingly apparent as local control failures and sudden accidents exposed its weaknesses. As a result, beginning in the 1980s, the field of automation and automatic control saw the rise of a wave of distributed control—DCS (Distributed Control System). This new control approach effectively divided control authority: simple loop control was handled by field-level control systems, while remote systems performed centralized functions such as signal setting and advanced control. From that point onward, computer technology and control technology became closely intertwined.

Yet the explosive progress of information technology, driven by Moore’s Law, did not bring automation control technology forward at the same pace. From the perspective of the IT industry, it would be almost unimaginable for dozens of companies to keep using essentially the same technology year after year, when several generations of technology can emerge, evolve, and disappear within just a decade or so.

Although automation control technology developed much more slowly than the IT industry, in the late 1980s and early 1990s the so-called fifth-generation control technology—fieldbus technology—also emerged in response to the broader trend toward openness brought about by the information industry.

In the era of the previous generation of distributed control systems, most manufacturers used closed systems and proprietary programming languages with their own unique code bases, which was not conducive to communication and exchange between systems. Fieldbus technology, by contrast, emphasized a shift from distributed control to field-level control. Data transmission adopted more advanced bus-based architectures. From then on, open protocols and the communication and exchange of information among devices and instruments from different manufacturers became the mainstream direction of automation technology.

However, the rise of fieldbus did not break the industry monopoly of major players. From that time on, leading automation manufacturers in the United States, Germany, and Japan established their own camps and built proprietary protocol standards within the broader international standards framework. Two major systems emerged: one was the Fieldbus Foundation system, represented mainly by major U.S. and Japanese companies and also known as the Foundation Fieldbus protocol system; the other was the Profibus protocol system, represented by German companies such as Siemens. Under these protocol standards, companies could communicate and exchange data, but only if they had a thorough understanding of each other’s interfaces and equipment. High barriers to data exchange and field management still remained. To avoid unnecessary trouble, users often preferred to purchase products from a single manufacturer or from manufacturers within the same protocol camp.

From that point on, the field of process control gradually moved toward a semi-open model. At the same time, within user sites, inside these automation devices, and within the manufacturers themselves, the rapid spread of large-scale integrated circuit chip technology—together with the rapid development of software technologies related to intelligent modeling and intelligent recognition—also led to the gradual adoption of a new form of control technology: intelligent control.

Traditional loop control and closed-loop control are based on given design parameters. Intelligent control, by contrast, first appeared in early aerospace and weapons systems. It emphasizes pattern recognition and the establishment of expert systems capable of continuously learning from previously collected information. This depends on the continued improvement of chip computing power and software technology. As intelligent control systems gained traction, control technology also began to merge rapidly with production systems. The automotive industry, represented by companies such as Toyota in Japan and Ford in the United States, began integrating computer control systems with production systems. The automotive industry was the first to introduce the flexible manufacturing system (FMS). China’s computer integrated manufacturing system (CIMS), developed over the past 20 years, is also representative of this process.

In fact, today’s automation control technology is continuing to advance rapidly along this very path, and its integration with the industrial internet is becoming increasingly evident. Many innovative new companies are emerging in this field and making major breakthrough innovations. For example, as chips improve in self-learning capability and as better models are developed, automated monitoring systems can significantly improve monitoring efficiency while reducing the computational burden on central computing systems. All of this is precisely part of the technological foundation of the industrial internet. The artificial neural networks and artificial intelligence systems now emerging are also branches of this new form of intelligent control.

IBM’s cognitive learning system Watson is aggregating global medical information and data, enabling it to surpass the learning capacity of any single human individual in the field of AI-assisted diagnosis and treatment.

Once such systems are applied in industry, they will greatly improve industrial process control, production management, and inventory management. In the future, AI-based control and management will challenge the traditional field of industrial automation. Even today, however, the technologies used in automation and control are still far from being deeply integrated with information technology. The fact that long-established control systems are still being marketed after many years reflects a basic reality: industrial systems evolve much more slowly than information technology, and over the past 20 years, IT companies and talent have not paid enough attention to the industrial sector.

With the rise of the industrial internet concept, companies seeking to rebuild their competitive advantage—especially Chinese companies that are currently in a catch-up phase—need to quickly bring internet-era technologies, software, and talent into the industrial field in order to trigger a new industrial revolution. Unfortunately, across China today, there are not many automation companies that possess both this mindset and the ability to act on it. That, of course, represents a tremendous opportunity, because the early movers will seize the commanding heights of this opportunity.