Business studies is a highly practical discipline, and the establishment of business schools aims to train professional management talent for enterprises. In the over a century-long history of global business schools, this original intent was maintained for several decades. Since the 1970s and 1980s, we have noticed a concerning trend: the growing disconnect between theory and practice, even to the extent of drifting apart. To return to the essence of business studies, business schools worldwide are trying various reformative ideas, hoping to normally and closely integrate theory with practice.

In 2018, Antai College of Economics and Management proposed the development strategy of "Interwoven Theory and Practice, Unity of Knowledge and Action," aimed at creating a business education ecosystem where academic and industry research complement each other. Industry research has become a bridge linking theory and practice. The Industry Research Institute at Shanghai Jiao Tong University was established accordingly, providing a platform for teachers and students to engage in industry practice. In the past five-plus years, the institute has formed over thirty research teams annually, covering major industry sectors vital to national welfare and people's livelihood, such as health, elderly care, finance, energy, manufacturing, services, distribution, and cultural creativity, making positive contributions to society, government, enterprises, teaching, and scientific research.

In 2023, I set an industry research goal for myself: to lead a team across China's lighthouse factories. (The lighthouse factory concept, proposed by the World Economic Forum in collaboration with McKinsey in 2018, selects the world's best manufacturing enterprises based on sustainable development, production capacity, agility, market response, customization, etc.) At the beginning of 2023, there were 132 lighthouse factories globally, 50 in China; by the end of the year, the global number increased by 21, with 12 more in China. Unfortunately, this research goal was more challenging than I anticipated, and we ultimately surveyed 30 benchmark enterprises in intelligent manufacturing, including 22 lighthouse factories, which I suppose is passable. Each research team was not small, consisting of more than ten Antai teachers, colleagues from the Industry Research Institute, alumni interested in intelligent manufacturing, and sometimes industry experts. The research spanned ten months, crossing 12 cities in six provinces and one municipality, including Jiangsu, Shandong, Anhui, Sichuan, Hunan, Guangdong, and Shanghai, covering industries like equipment manufacturing, home appliances, automobiles, and fast-moving consumer goods.

This article reflects some thoughts from the past year's enterprise research, many of which benefitted from discussions with team members. I take this opportunity to thank them and the enterprises that warmly received us and patiently answered our questions.

I. The Development Path of Productivity

The degree of human civilization depends on the level of productivity development; the more advanced the productivity, the higher the civilization level.

The development of productivity can be attributed to advancements in two aspects: "physical power" and "brain power." Just as an individual's ability is determined by their physical and brain power, the level of societal productivity depends on society's "physical" and "brain" power levels. For instance, the tools and equipment we use represent our society's "physical" power level, an extension of our "hands" and "legs," generally stemming from technological progress, representing the contributions of scientists and engineers. Meanwhile, society's "brain power" refers to our adopted work or production methods, an extension of the "brain." In essence, societal "brain power" is about management or governance, i.e., the effective utilization of resources, determining whether we work smartly or hard and our efficiency. The advancement of societal brain power comes from management innovation, primarily originating from the ingenious ideas of managers and the summarization and sublimation by management scholars, and may also benefit from technological progress.

This dualistic view of productivity, comprising "physical" and "brain" power, helps us clarify the development course of human productivity. Throughout history, the development of productivity has unfolded along these two dimensions. The steam engine of the First Industrial Revolution and the electricity of the Second Industrial Revolution propelled human society into the mechanical and electrical eras, significantly enhancing societal physical power. However, we must not forget that along with these two industrial revolutions, societal brain power also made great strides. For example, Ford's assembly line and Sloan's divisional management method, these epoch-making management innovations, significantly increased production efficiency and societal satisfaction. The Third Industrial Revolution's computers and the internet ushered human society into the information age, and these technological changes simultaneously promoted the development of societal physical and brain power, as computers made devices more flexible and production lines more adaptable. At the same time, computer technology brought a series of industrial software, greatly enhancing management efficiency and strengthening societal brain power. Interestingly, parallel to the Third Industrial Revolution was another profound management transformation: the introduction of the Toyota Production System, later known as the Lean Production System. This profound and comprehensive management system quickly spread globally, changing the world by significantly reducing waste and increasing efficiency, marking another significant advancement in societal brain power. Note that the introduction of lean thinking was not greatly related to computer/internet technology; it was continuously explored and summarized by generations of Toyota managers, contributing significantly to advancing human productivity. Lastly, the Fourth Industrial Revolution, as often discussed, is the intelligent era. As the name suggests, this will be a significant explosion of societal brain power, where the new generation of digital technology and artificial intelligence will directly empower brain power, significantly enhancing management levels and improving resource utilization efficiency. This was continually confirmed in our research on intelligent manufacturing enterprises.

In summary, each industrial revolution begins with a core technology, which then triggers widespread social changes and rapidly enhances human productivity. Although productivity enhancement usually includes progress in both "physical" and "brain" power, we notice that discussions about industrial revolutions typically only focus on the changes in societal physical power triggered by technology, paying insufficient attention to the development of societal brain power. This might be excusable for the First and Second Industrial Revolutions, as the steam engine and electricity did not directly affect the development of societal brain power, although societal brain power was also progressing (like Ford and Sloan's innovations). This situation has greatly changed in the Third and Fourth Industrial Revolutions, as the involved core technologies directly promoted the advancement of societal brain power, especially the core technology of the Fourth Industrial Revolution - artificial intelligence - which will create humanity's greatest "external brain." Therefore, there will be great attention to the development of societal brain power. In other words, management innovation will become a main theme of this industrial revolution, and intelligent manufacturing will be a splendid sight in this wave of global transformation.

Societal Brain Power: Its Manifestation Across Industries

In manufacturing, societal brain power is manifested as the level of intelligence in factories. Simply put, it's about the "smartness" of a factory's operations. The advent of new digital technologies has made intelligent factories possible. If we use human bodily functions as a metaphor, an intelligent factory must possess senses, nerves, and a brain, which correspond to information gathering, transmission, and processing capabilities. These functions can be found in the treasure trove of modern digital technology, such as sensors, the internet/IoT, and data/supercomputing centers.

Perception System: This system gathers and interprets physical (optical, positional, weight, speed, pressure, etc.) and chemical (humidity, pH, ion concentration, electrochemical gases, etc.) information involved in the production process. It provides real-time data and feedback for the manufacturing system, enabling the timely identification and resolution of issues, thereby improving production efficiency and quality.

Information Transmission: Relies on industrial Ethernet, fieldbus, wireless communication, and other communication technologies to enable real-time and efficient information transfer between devices, sensors, and controllers, achieving coordinated operation. Intelligent manufacturing systems also need to transmit information with external systems, such as ERP systems and supply chain management systems, to automate processes like production planning, material procurement, and inventory management.

Decision System: Utilizes advanced solving techniques to analyze various data and information in production/distribution/sales processes for optimal decision-making. A robust decision system can help enterprises reduce costs, improve efficiency, enhance product quality, better serve the market, and ultimately enhance competitiveness.

Execution System: This is a software system used for real-time monitoring, immediate coordination, and timely management of the manufacturing process. Closely associated with production sites and support, its main functions include material distribution, production scheduling, equipment management, quality control, and process management. The execution system precisely implements decisions through corresponding industrial software, control systems, and hardware, achieving closed-loop control over the perception, transmission, decision, and execution of the production process.

It's important to note that the cost of these digital technologies is not low, with million-level investments often being just a basic threshold. Therefore, enterprises need to focus on specific problems, carefully analyze input and output, and avoid digitizing for the sake of digitization. Next, we will introduce how intelligent manufacturing enterprises use digital technology to enhance management and achieve operational benefits.

II. The Essence of Digital Transformation is Management Innovation

The essence of management is to identify and solve problems. The core value of digital transformation lies in better identifying and solving problems.

In the 1980s, the "Toyota Production System (TPS)" proposed by Toyota, later summarized as "Lean Production," is a very comprehensive management system. It details various aspects from goals to methods, systems, and corporate culture. In my view, Lean Production truly captures the essence of management. For instance, its understanding of "waste" is profound and systematic. If an employee's process improvement idea is not timely valued and implemented, it's a form of waste, as is a lack of perfect supply-demand matching, a prominent manifestation of which is inventory, the core idea behind Just-In-Time (JIT) production, etc. Lean Production views all forms of waste as the "number one enemy," and the underlying issues must be addressed. It also emphasizes timely problem detection, such as the "andon" system, a device that leaves no hiding place for any issue. Once a problem arises, relevant personnel are immediately informed and take action, as any delay is waste. Of course, Lean Production also provides many mechanisms and even cultural arrangements for problem-solving, like "everyone is a scientist," "continuously asking five whys," "problem-solving teams," "lifetime employment relationships," and more. Thus, I believe Lean Production excellently covers the three major articles of "problems," "detection," and "resolution."

In my view, Japan's rise in the 1980s was largely due to the rise of its manufacturing industry, driven by the Toyota Production System or Lean Production. This epoch-making management innovation significantly enhanced the aforementioned "societal brain power," injecting a strong force into the enhancement of social productivity. Those who visited Toyota factories in that era remember that the hardware equipment was very ordinary, even rudimentary. Yet, such conditions could produce such strong productivity, the reason being management "brain power." This is another powerful testament to the previously mentioned dual theory of productivity, where both "societal physical power" and "societal brain power" are essential sources of productivity.

In today's digital age, the essence of management hasn't changed; it's still about identifying and solving problems. Only now, our means of identifying problems have become more advanced, and our ability to solve them has strengthened, enabling us to address an ever-increasing number of issues.

l  Advancing Means of Problem Identification

Human methods of identifying problems are sophisticated. We have six senses: vision, hearing, touch, taste, smell, and intuition, which gather information from all directions. This information is then transmitted to the brain via neural networks. Of course, the human sensory system has its limitations, and there remains a vast unknown space. However, these shortcomings are being addressed by modern digital technologies. Various sensors can capture information such as temperature, humidity, vibration, images, sound, and language. The Internet functions like a neural network, transmitting collected information to data/monitoring/decision centers. During factory surveys, we witnessed numerous applications of sensory technologies. For example, the Bosch Wuxi factory uses machine vision technology for quality inspection. By continuously taking micron-level photographs and using high-precision AI judgment, it compares part photos with standard images to conduct real-time intelligent diagnostics. This resolves the fatigue and inaccuracy of manual visual inspection and shifts from post-production batch sampling to 100% real-time monitoring during production. Another interesting application at the Bosch factory is the "Intelligent Fingerprint" anti-counterfeit label scheme. Steel components produce random patterns during grinding and polishing, which can be captured and recorded by machine vision technology. Combined with a system-generated encrypted QR code, an "Intelligent Fingerprint" is created. Compared to traditional anti-counterfeiting systems, this "Intelligent Fingerprint" offers advantages like low cost, high security, identifiability, and traceability. Additionally, sensory technology is also used in health management. Foxconn in Chengdu created an online health management platform for employees, using a self-developed device, "Xiao Yu Jing Ling," to measure heart rate, blood oxygen, uric acid, and other indicators. Through big data analysis and other technologies, employees receive personalized health advice and preventive measures. Maintaining physiological and psychological health helps employees effectively manage work stress, fostering their enthusiasm and creativity. All these not only maintain production efficiency and stability, reducing safety risks, but also reflect corporate social responsibility. During the pandemic in the first half of 2022, Foxconn's Chengdu Intelligent Park, with 125,000 employees, achieved the remarkable feat of zero infections.

l  Increasing Capability to Solve Problems

From a business management perspective, our problem-solving capability is mainly manifested in two aspects: the complexity of the problem and the degree of decision optimization. The higher the complexity and optimization level, the stronger the capability. Driven by digital technology, enterprises can solve increasingly complex problems, achieving closer to optimal solutions. This is primarily due to the rapid development of computing power and algorithms (such as high-end chips, supercomputing centers, solvers, software systems, etc.), which enable us to process massive data quickly and find the best (or approximate) solutions in vast variable spaces. At Bosch's Changsha factory, we saw a factory-wide production energy management system. Its goal is to minimize energy consumption while ensuring production and order fulfillment. The factory has over 100 production lines with hundreds of daily customer demands, making scheduling incredibly challenging, especially with the added goal of reducing energy consumption. The factory developed a production energy consumption prediction model, combining various business data (customer demand, equipment energy information, etc.) and environmental data (weather conditions, temperature, humidity, etc.), to provide rolling 7-day line-level production scheduling and energy predictions (with less than 3.2% error). Simultaneously, the system offers optimized shutdown management and precise energy consumption anomaly monitoring, integrating energy saving and emission reduction into every aspect of production manufacturing, such as prioritizing the use of new, less energy-consuming equipment when idle. This management approach has significantly contributed to the enterprise's ability to cope with complex energy environments, like sudden power restrictions in summer. Another example is Tsingtao Beer's use of digital technology to empower product research and development. Beer contains a high number of flavor substances, over 500 types (compared to 200 in baijiu), and different ratios result in different tastes. Imaginably, endless variations of beer flavors can be designed through different combinations. To better serve consumers' personalized demands, Tsingtao uses flavor map decoding technology, digitalizing taste, flavor substances, and process technology. For instance, consumer taste indices include alcohol-to-ester ratio, malt aroma, hop aroma, fruit aroma, floral aroma, etc.; typical flavor substances include alcohols, esters, organic acids, etc.; process technology includes yeast strains, raw material standards, formula standards, and fermentation process standards. The digital-driven flavor map decoding technology has tripled Tsingtao's R&D efficiency, refined tastes, and rapidly met consumers' personalized demands, achieving market leadership.

l  Increasing Diversity of Solvable Problems

Digital technology has enhanced our ability to identify and solve problems, thereby increasing the variety of problems we can address. Previously, we often dealt with issues that had already arisen, but now we can prevent problems before they even emerge. As is widely known, every problem undergoes a development and maturation process. In our factory research, we observed some enterprises already tackling these "immature problems." In other words, management is moving upstream in the "problem chain," eliminating the root causes of problems and altering their development trajectory, akin to the traditional Chinese medicine philosophy of "preventing disease before it occurs." Bosch's Changsha factory's predictive maintenance of equipment is an excellent example. Ninety percent of the production lines are interconnected, and intelligent sensors and visual devices capture a range of process parameters like vibration, torque, voltage, and current in real time, understanding equipment dynamics at the millisecond level. The data team uses deep learning algorithms to build and continuously strengthen models through self-learning and training. This system automatically monitors equipment anomalies and their causes, predicts the health of the equipment within 24 hours, and issues warnings, aiding maintenance and production staff in preparing and responding quickly. This reduces the incidence of emergency maintenance orders and ensures production stability. Of course, the diversity of problems is also reflected in many other aspects. For example, Midea's "one stock" practice reformed the traditional distribution channel by "disintermediation," separating "goods rights" and "sales rights." Agents and distributors only have "sales rights," while "goods rights" or inventory are centralized in the group's central warehouse, which completes distribution, achieving "one stock" across all channels. Here, management addresses the integration of supply chain upstream and downstream, including the redesign and management of information and goods flow, supported by full-link networking, large-scale market information capture, and a powerful demand forecasting model. Another example is Kute Intelligence, a Shandong-based custom suit enterprise. To effectively address the challenges of large-scale, personalized custom production, they built a C2M platform, breaking down customer orders into specific work tasks, then directly transmitting task information to every workstation on the production line, eliminating hierarchical structures and increasing efficiency. Employees, no longer having a "boss," directly follow instructions from the information platform, enhancing their sense of happiness. Similarly skilled employees form a "cell unit organization" that self-organizes and coordinates, reallocating saved management costs to employee benefits, further enhancing their sense of achievement. This is an example of optimized organizational structure. Driven by digital technology, new organizational structures become possible.

III. Thoughts and Outlook

Future Trends in Intelligent Manufacturing: Manufacturing is a critical matter concerning national welfare and people's livelihood, playing a pivotal role in wealth creation and employment provision. Intelligent manufacturing represents a significant opportunity for the comprehensive improvement of the manufacturing industry. Over the past year, we visited several excellent manufacturing enterprises, which are models of industry practice, setting benchmarks for digital transformation. Although digital technology is rapidly evolving, the essence of management remains unchanged: continuously identifying and solving problems. The sophistication of management can be gauged by how advanced the means of problem identification are, how strong the problem-solving capability is, and what specific problems are addressed. We also found that the concept of Lean Production is far from outdated, with excellent enterprises considering it the foundation of intelligent manufacturing. Digital technology, indeed, gives wings to Lean Production, elevating a traditional (yet timeless) management concept to new heights. In the future, I believe that means of problem identification will continually improve, problem-solving capabilities will constantly enhance, and the boundary of solvable problems will keep expanding. For instance, digital twin technology, as the name suggests, opens a parallel virtual space to the real world. In this virtual world, we can conduct various simulations, foreseeing the "future" and thus acquiring the ability to "prevent disease before it occurs." With such digital technology, we can aspire to a "worry-free factory," where any problem can be nipped in the bud. Additionally, digital technology, especially the rapid development of artificial intelligence, will greatly enhance human "brain power," leading to leaps in information collection, integration, and decision optimization. In short, human productivity will receive boundless vitality from digitally empowered brain power.

The Importance of Manufacturing Ecosystem Construction: The 30 enterprises we visited last year are industry leaders, but this doesn't mean we only focus on the top of the pyramid. On the contrary, I believe the healthy development of the entire manufacturing ecosystem is crucial – there can be no peak without a base. A common saying is that competition between enterprises is actually competition between supply chains, underscoring the importance of the supply chain. I think a better phrase is that competition between enterprises is, in fact, competition between industrial ecosystems. An enterprise's competitiveness depends not only on its own actions but also significantly on the health of its industrial ecosystem. One important reason China has the most lighthouse factories globally is the wealth accumulated from years of reform and opening up, i.e., a quality industrial ecosystem. We must protect this ecosystem well. We can learn from the changes in the American industrial ecosystem. As is well known, in the 1980s and 1990s, American manufacturing began an outsourcing wave, moving labor-intensive factories to countries with lower labor costs, especially China. This wave severely impacted the U.S. manufacturing ecosystem, causing widespread unemployment and many subsequent political and social issues, and even laying the groundwork for the current tension between China and the U.S. Our industrial policy should not only focus on "high-end" sectors but also create a vibrant industrial ecosystem, requiring "diversity of species." A healthy industrial ecosystem is an important step toward achieving Chinese-style modernization, especially for a country with a vast population.

Antai's Ongoing Industry Research: As previously mentioned, intelligent manufacturing is an important opportunity for the comprehensive enhancement of China's manufacturing industry. Rapidly evolving digital technology offers many new opportunities for enterprises, from product innovation to management and business model innovation. We must seize the opportunity of the Fourth Industrial Revolution, allowing digital technology to benefit all enterprises and continuously enhance their competitiveness. Our enterprise research has just begun. In the future, we will expand the scope of our research, closely monitor the development of industrial ecosystems, analyze successes and failures, study the digital transformation paths of different types of enterprises, and hopefully eventually summarize a management theory with Chinese characteristics, contributing Chinese wisdom to the global manufacturing industry's development.