《1 Engineering research fronts》

1 Engineering research fronts

《1.1 Trends in Top 10 engineering research fronts》

1.1 Trends in Top 10 engineering research fronts

In the field of engineering management, the global Top 10 engineering research frontiers this year are “research on industrial digital transformation empowered by industrial internet platform”, “research on global supply chain security risk management in the digital age”, “research on big data governance methods in artificial intelligence scenarios”, “research on the theory and method of accurate construction and evolution of digital twin model”, “research on sustainable transportation system under carbon peaking and carbon neutrality strategy”, “formation mechanism, evolutionary law and governance policy of global public health crisis”, “medium- and long-term sustainable supply paths and policies for basic and typical strategic resources”, “group consensus mechanism under social network”, “big data-based financial risk assessment”, and “research on intelligent operation and maintenance management of infrastructures from the perspective of socio-technical system theory”. The core papers published are listed in Table 1.1.1 and Table 1.1.2. Among them, “research on industrial digital transformation empowered by industrial internet platform”, “research on global supply chain security risk management in the digital age”, “research on big data governance methods in artificial intelligence scenarios”, and “research on the theory and method of accurate construction and evolution of digital twin model” are the key fronts for interpretation, and their current development trends and future trends will be explained in detail later.

(1)  Research on industrial digital transformation empowered by industrial internet platform

The industrial Internet platform is the carrier of scenario- driven deep integration of new generation information technology and manufacturing industry and is the key infrastructure for digital technology and data elements to collaboratively drive the digitalized, networked and

《Table 1.1.1》

Table 1.1.1 Top 10 engineering research fronts in engineering management

《Table 1.1.2》

Table 1.1.2 Annual number of core papers published for the Top 10 engineering research fronts in engineering management

intelligent transformation and upgrading of the industry. As the digital economy and a new round of technological revolution are evolving in depth on a global scale, industrial Internet platforms such as Amazon AWS, GE Predix, Siemens MindSphere, Haier COSMOPlat, Ali supET, Sany ROOTCLOUD, yonyou ERP, Huawei FusionPlant, etc. have become the key engines, important ways and new carriers dynamic capability and accelerating the digital transformation of industry. The industrial Internet platform, through a comprehensive digital empowerment platform, realizes the comprehensive interconnection of humans, machines, and things and the comprehensive linkage of the whole industrial chain, the whole value chain, the whole innovation chain, and the whole staff and the whole elements, and promotes the deep integration of the digital economy and the real economy. It can meet the complex comprehensive and personalized demands of multiple industrial scenarios with higher efficiency, lower cost, and dynamism, and promote the formation of a new industrial manufacturing and digital and intelligent service system. The research on industrial digital transformation empowered by industrial internet platform mainly focuses on the technical architecture of industrial Internet platform, the process mechanism of empowering industrial digital transformation, the dynamic capability of industrial digitalization, the ecological construction of integration and innovation of large, medium and small enterprises, and the governance of industrial Internet platform. The development of new generation digital technologies, such as satellite internet, federal learning, privacy-enhanced computing, Web 3.0, and Metaverse, and the demand trend of green and low carbon development make industry cloud-native platform, industrial Metaverse, scenario-driven innovation and so on become the future trend of the research on industrial digital transformation empowered by industrial internet platform.

(2)  Research on global supply chain security risk management in the digital age

In the new round of technological revolution and industrial transformation, digital technology is driving great changes in the global supply chain industry, and the supply chain has ushered in a digital, networked, and intelligent “digital change”. Although digital technology innovation is gradually pushing the expectation of global industrial services to its peak, the ensuing global supply chain security issue has also become increasingly prominent. On the one hand, in the critical period of intensifying international trade friction, increasing uncertainties, and high-quality transformation of industries, the supply chain risk problems brought by weak key technologies are gradually exposed. From the root, the problem of key technology “bottleneck” comes from the imperfect layout of the key industrial chain. Especially under the current trend of reverse globalization and active decoupling of supply chains in developed countries, how to ensure the supply chain security of global high-tech industries has become an increasingly urgent issue. On the other hand, although the emergence of the digital economy has effectively promoted the development of supply chain transparency and wisdom, the information leakage problem brought by data integration has been frequent, the digital divide between countries has been expanding, and the struggle for digital sovereignty has entered into a white heat, which has affected the rapid development of global smart supply chain. In recent academic research, the key research directions are the industrial chain layout method of key technologies in the supply chain, supply chain security issues in high- tech industries under the intensified economic competition of reverse globalization, and supply chain data leakage and information security threats caused by data integration.

(3)   Research on big data governance methods in artificial intelligence scenarios

Big data governance is a set of management behaviors involving the use of big data in an organization. It contains two definitions, one is “governance of big data”, that is, to use a certain method or form for continuous governance of big data itself, such as data source import, cleaning, and processing, data normalization, data storage, data computing, data service application and so on to improve the quality and value of data, which is conducive to the subsequent use of big data. The other is “relying on big data for governance”, that is, to use big data, cloud computing, artificial intelligence (AI), and other advanced technologies to achieve intelligence of process regulation, efficiency improvement, and social governance means. Firstly, the key technologies of big data governance under artificial intelligence scenarios mainly include structured data processing, data quality assessment and data cleaning, data normalization, data fusion and extraction, data organization, and data sharing, and other whole process service links. Secondly, by constructing big data mapping knowledge domain, it realizes understanding data, explaining phenomena and knowledge reasoning, discovering deep relationships, and realizing intelligent search and intelligent interaction. Thirdly, big data governance under the AI scenario faces the problem of security controllability. There is a need to solve data security and algorithm security problems such as privacy leakage, data rights validation, algorithm bias, and technology abuse, to promote social wisdom governance and intelligent industrial transformation. Finally, the typical AI application scenarios of big data governance are applied to more complex and high-value scenarios such as finance, healthcare, urban management, public opinion monitoring and so on for solving management needs such as business decision-making, resource allocation, process optimization, operation and maintenance assurance, and risk prevention and control.

(4)    Research on the theory and method of accurate construction and evolution of digital twin model

Digital twin models are digital representations of real-world entities or systems that can be used to understand, predict, optimize, and control physical-world entities or systems. Therefore, the construction of the digital twin model is the basis for model-driven implementation. The construction of the digital twin model is the digital modeling of the properties, methods, and behaviors of physical entities and processes in digital space. The model construction can be multidimensional in “geometry-physics-behavior-rules”, or multi-domain in mechanical, electrical, hydraulic, and other fields. The digital twin model should cover multi-dimensional and multi-domain models to achieve a comprehensive and realistic characterization and description of physical objects. For building models of relatively complex objects, it is necessary to solve the problem of how to assemble and fuse simple models to form complex models. To ensure the correctness and validity of the digital twin model, the model needs to be validated and evolved to ensure that the model describes and characterizes the state or characteristics of the physical object correctly, i.e., to ensure the virtual-real consistency of the model. Therefore, the main research direction can focus on the six stages of the digital twin model, i.e., “construction-assembly-fusion-validation-correction- management”, for carrying out the accurate construction of the multi-dimensional/multi-domain digital twin model, the assembly and fusion of all-elements/multi-scale twin digital model, the validation and correction of virtual-real consistency of digital twin model, and the interactive iteration and dynamic evolution of digital twin model. The digital twin must accurately represent the physical system in its current state, which requires the digital twin model to reflect the changes and updates of the physical system through fast and reliable accuracy. In the future, the construction and evolution of digital twin models will definitely develop in the direction of improving modeling efficiency and accuracy.

(5)   Research on sustainable transportation system under carbon peaking and carbon neutrality strategy

The process of promoting sustainable transportation systems is a search for more scientific energy structures and energy use mechanisms, lower emissions, and easier ways and technologies for system emissions management or self- absorption and solidification, including the development and implementation of policies, standards and laws, and regulations.

In 1992, China became one of the first parties to sign the United Nations Framework Convention on Climate Change, and in 2002, the Chinese government ratified the Kyoto Protocol. With active promotion by China, the world reached the Paris Agreement to address climate change in 2015. In 2016, China took the lead in signing the Agreement. In September 2020, President Xi Jinping announced at the 75th UN General Assembly that China strives to peak its CO2 emissions by 2030 and work towards carbon neutrality by 2060. This carbon peaking and carbon neutrality strategy proposes a new direction for the research and advancement of sustainable transportation systems.

Concerning sustainable transportation systems, research at the environmental level has focused on three main areas, including the reduction of pollution emissions, the shifting of emissions, and the management of emissions. From the perspective of emission reduction, from land transportation to aviation, to water transportation, cleaner energy is promoted and how to overcome technical or cost constraints is studied, allowing cleaner energy more widely available than traditional energy. Another path to emission reduction is to continuously optimize the use of vehicles and equipment, the construction, maintenance, and use of facilities, and the allocation of time, space, and related resources. Thirdly, reduce energy consumption through technological upgrading. In the context of the carbon peaking and carbon neutrality strategy, the research has gradually shifted from regional air quality issues to global climate change issues, and the focus is increasingly put on carbon emissions. From the perspective of shifting emissions, both the promotion of new energy vehicles and the setting up of marine pollution emission control zones fall within the scope of the research. The research in this field often does not start with reducing carbon emissions but focuses on reducing the emission of harmful gases in high-density population areas. At present, the research on pollutant management mainly covers harmful gases, and the research on carbon neutrality is still emerging.

(6)  Formation mechanism, evolutionary law and governance policy of global public health crisis

A global public health crisis is an event induced by a public health emergency, which brings a serious impact (or even causes a shutdown) on the normal operation of social systems in most countries around the world, and has a serious impact on international economic and trade exchanges, national security, etc. The main research directions include:

1)   Research on the mechanism of formation and evolution of global public health crises. This direction focuses on the evolutionary mechanism of local health events into global public health crises, the mechanism of the coupling of complex systems such as economy, transportation, and information on the evolutionary process, and the failure mechanism of existing monitoring and warning and risk governance systems, and the vulnerability analysis of different populations in crises.

2)   Research on the governance strategy of global public health crises. This direction focuses on scarcity management, capacity expansion, supply chain failure, and other optimized resource allocation and institutional mechanism design issues of catastrophic peak demand scenarios, international participation, and international cooperation and coordination mechanisms in the governance of global public health crisis, the laws, and mechanisms of research and decision making of multiple subjects in typical scenarios, and the analysis of the impact of the implementation of typical policy tools on the interventions with the mentality and behavior of multiple subjects.

3)  Research on learning, change, and resilience enhancement after the global public health crisis. This direction can focus on the recovery and learning mechanisms of public and private organizations, institutions, and societies after crises, the impact mechanisms of the combination of fiscal and taxation policy tools on the recovery of individuals, enterprises, and industrial chains, the laws of change and adaptation of multiple subjects after crises, the top-level design, and model reconstruction of the response system, and the strategies of resilient governance enhancement.

The probability of global public health crises is low, and governments and other subjects lack long-term attention to system building and capacity reserve. How to use the COVID-19 epidemic crisis as an opportunity to promote research, system building and capacity reserve will be a topic worthy of attention in the future.

(7)   Medium- and long-term sustainable supply paths and policies for basic and typical strategic resources

Basic and typical strategic resources generally refer to resources that are related to global livelihoods and dominate the resource system, and their sustainable supply depends on factors such as resource endowment, supply and demand, economic development, reserve effectiveness, and international material availability. Traditionally, resources such as energy, minerals, water, and land are the most typical basic and strategic resources. However, with the progress of science and technology and global economic development, data resources, highly skilled professional labor, and carbon emission rights are also included in the category of basic and typical strategic resources. At present, systematic research has been conducted on the medium- and long-term sustainable supply paths and policies for basic and typical strategic resources, and an analysis framework of sustainable supply paths and policies has been constructed from both production and consumption sides, focusing on the life cycle measurement and management of basic and typical strategic resources, improvement of resource utilization efficiency, environmental health and waste recycling, supply chain security risk assessment, government regulation and optimal allocation of resources, and sustainable supply performance evaluation. Based on technical methods such as systems thinking and artificial intelligence, exploring a set of thinking paradigms, practice paradigms, and research paradigms for sustainable supply and management policies of basic and typical strategic resources from a medium- and long- term perspective will become the focus of future research breakthroughs.Key scientific issues such as the lifecycle benefits and cost optimization of basic and typical strategic resources, sustainable supply chain modeling and governance paths, green technology innovation and circular economy, global climate change mitigation and sustainable resource supply, optimal allocation of resources from a blockchain perspective, resource security and policy simulation analysis, artificial intelligence and sustainable management decision system are frontier research hotspots at the intersection of multiple disciplines.

(8) Group consensus mechanism under social network

Group consensus seeks to reconcile the conflicting views of decision makers and to find a group decision solution that unifies the majority opinion. Traditional group consensus models assume that decision-makers are independent of each other, but in fact, decision-makers are in a social network, and their mutual relationship with each other and the structural characteristics of the corresponding network are important factors that affect group consensus.

In recent years, the widespread popularity of social media has accelerated the interaction and information transfer among decision-making individuals, and group consensus is facing more and more conflicts and challenges, such as the expansion of decision-making group size, diversity of interest groups, differences in individual preferences, and complexity of decision problems. Social networks are used to accurately characterize the interrelationships among decision makers (interest relations, trust relations, conflict relations, emotional relations, etc.), providing a new research perspective for the development of group consensus mechanisms. In the context of big data, the introduction of social network analysis methods into group decision-making scenarios can be widely used in scenarios such as emergency decision-making of major public opinion events and automatic discovery of social hotspots and exceptional events.

Currently, the main research directions of group consensus mechanisms under social networks include the research on group consensus models in dynamic social network environments, consensus compensation and non-cooperative behaviors in social network environments, and malicious manipulation behaviors in social network environments. In addition, with the development of artificial intelligence and Metaverse technologies, there are social robots as well as online virtual humans in addition to real humans on social networks. How these new social relationships (e.g., human– machine interaction and machine–machine interaction) affect the group consensus process urgently needs to be studied in depth.

(9) Big data-based financial risk assessment

Data are numerical characteristics or information obtained through observation. With the development of digital technology and its wide application in the financial field, financial data with diverse types, massive heterogeneity, complex relationships, low-value density, high noise, and non-normal characteristics continue to emerge at a high speed, forming a wave of financial big data. In this context, the research and application of financial technology in financial risk monitoring, assessment, and early warning are becoming more and more in-depth, which enables and enhances efficiency and also leads to the emergence of new and potentially harmful financial risk problems. Financial risks are characterized by complexity, concealment, cross- domain, transmission, and dynamics, which make them difficult to characterize, cognize, identify, track, model, reason, and evaluate. The current monitoring, evaluation, and early warning of financial risks are based on the analysis and processing of massive financial data, and the identification and prediction of risks by establishing algorithmic models. The research mainly focuses on the security protection, safe storage, safe transmission, real-time processing and analysis and sharing circulation of massive financial data, the rapid identification methods of financial risks based on graph data, text data, stream data and other non-structural data, research on the prevention and early warning of systemic financial risks using graphical models and statistical models, research on the use of data science for financial credit risk assessment, etc. Future research trends include the research on financial data sharing and traceability methods under the division of management system, research on autonomous and controlled financial data intelligent twin and sensitive feature masking technology, research on financial twin data privacy risk and data quality assessment technology, research on financial data twin security environment construction and data deposition technology, research on financial data twin-based approximate query processing and next-generation financial database test benchmark building technology, research on financial risk behavior characterization, cognition and financial network risk transmission, modeling and evaluation, research on big data-driven technology for rapid risk identification and analysis of fintech products, construction of a large-scale financial risk awareness mapping system through financial big data, etc.

(10)   Research on intelligent operation and maintenance management of infrastructures from the perspective of socio- technical system theory

Modern infrastructure systems, with engineering facilities as carriers, rely on complex technologies and equipment to perform their functions and provide important support for social production and residents’ lives. On the other hand, its operation and maintenance management involves many aspects such as operation, maintenance, and consumer services, and is influenced by the decisions and behaviors of various stakeholders such as technicians, management teams, and end users, as well as the constraints of external conditions such as economic, social and environmental conditions. Therefore, infrastructure has typical socio- technical system characteristics. With the continuous development of infrastructure, the interactive coupling between its social and technical components becomes more and more frequent and complex, which has an increasingly critical impact on the effectiveness and safety of infrastructure. In recent years, domestic and foreign scholars have made significant progress in such hot research directions as socio- technical system modeling and simulation of infrastructure, operation and maintenance scenario construction and derivation of infrastructure and stakeholder interaction, infrastructure operation and maintenance optimization based on human factors engineering and behavioral decision theory, infrastructure security and resilience management based on information-physical-social convergence system, the pathways and management strategies for the role of infrastructure in socio-economic development and security, and policies and institutional mechanisms for green and low- carbon infrastructure development. In the context of the rapid development of information technology such as the Internet of Things, cloud computing, and artificial intelligence, how to further promote the paradigm, methods, and technological innovation of infrastructure operation and maintenance management based on the multi-dimensional integration of physical-social-information, optimize the interaction mode and efficiency of the technical components and social components of infrastructure, and realize the transformation of intelligent and low-carbon infrastructure and the continuous improvement of efficiency and security level, is expected to become an important research hotspot in the future.

《1.2 Interpretations for four key engineering research fronts》

1.2 Interpretations for four key engineering research fronts

1.2.1 Research on industrial digital transformation empowered by industrial internet platform

The focus of the industrial digital transformation empowered by the industrial Internet platform is to study how the

industrial Internet platform can accelerate the process mechanism of industrial digitalization, networking, and intelligent transformation and the ecological construction of integration and innovation of large and medium-sized enterprises and their effective governance by creating a comprehensive service platform of digital technology, data elements and intelligent manufacturing with deep collaboration, targeting the needs of multiple industrial application scenarios. After GE proposed the concept of the Industrial Internet in 2012, domestic and foreign policies and academic fields gradually reached a consensus that the Industrial Internet is the key new engine to drive the transformation of the manufacturing industry from the supply-side perspective after the consumer Internet. Since then, GE, together with IBM, Cisco, Intel, and other leading international enterprises, has established the Industrial Internet Consortium (IIC), and the industrial Internet platform has shown a global trend of spurt. In China, based on “Made in China 2025”, more and more attention has been paid to the traction role of the industrial Internet in further promoting the integration of informatization and industrialization and the digital transformation of industry. In 2021, the national industrial Internet platform exceeded 600, with more than 100 with certain industry and regional influence. In 2021 and 2022, China’s Ministry of Industry and Information Technology issued The 14th Five-Year Plan for Deeply Integrated Development of Informatization and Industrialization and Industrial Internet Innovation and Development Action Plan (2021–2023), both of which focus on the industrial Internet platform as a key direction. Especially under the trend that the risk of blockage and breakage in the global industrial chain and supply chain intensifies and the digital transformation of the manufacturing industry enters deep water, the industrial digital transformation and upgrading empowered by the industrial Internet platform become a more important and urgent issue.

In recent years, scholars at home and abroad have focused on topics such as the technical architecture of industrial Internet platforms, the process mechanism of empowering industrial digital transformation, the dynamic capability of industrial digitalization, the ecological construction of integration and innovation of large, medium and small enterprises, and the governance of industrial Internet platform. Among them, the research on the technical architecture of industrial Internet platforms focuses on different technical architectures such as IaaS, PaaS, SaaS, and the emerging AIaaS, breakthroughs in key common digital core technologies, localization of platform software and its differentiated business model to empower industrial digitalization. The research on the process mechanism of empowering industrial digital transformation focuses on the mechanism of organic integration of digital technologies, data elements, and platform modules and the application of multiple industrial scenarios. The research on the dynamic capability of industrial digitalization focuses on how the leading enterprises can realize their transformation and empower the all-around transformation of the industrial chain and value chain through the construction of a digital technology capability system, digital management capability system, and digital scenario application capability system. The research on the ecological construction of integration and innovation of large, medium, and small enterprises focuses on how to support small and medium enterprises lacking data elements and key digital technologies to participate in the ecological construction of industrial Internet platforms and accelerate digital transformation, to ensure the safe and resilient development of the industrial chain and supply chain. The research on the governance of industrial Internet platforms focuses on data privacy, platform monopoly, platform interconnection, and governance standard system.

Looking ahead, with the emergence of a new generation of digital technologies and the acceleration of the co-evolution of industrial Internet platforms and industrial digital transformation, industry cloud-native platforms, industrial Metaverse, and scenario-driven innovation become research directions worthy of attention and forward-looking layout.

(1) Industry cloud-native platform

Industry cloud-native is an emerging concept in the field of cloud computing and industrial Internet, a new thinking and innovative model for industrial Internet platform construction and service application, and a key infrastructure for platform services. It can realize agile application development, iterative efficiency improvement, and delivery speed of industrial Internet platform for industry personalized needs. Industry cloud-native platform is still in the rapid emergence and exploration stage, and further attention needs to be paid to the as-a-service delivery model, continuous software-defined delivery, distributed enterprises based on industry cloud- native platform, and the construction of standardized and highly automated cloud services and intelligent decision- making systems.

(2)  Industrial Metaverse

Industrial Metaverse is the application of Metaverse-related technology and software in the industrial field led by a new generation of artificial intelligence technology. It can realize the virtual-real mapping of physical space, virtual space, and cyberspace in the industrial field as well as the interactive integration, thus realizing the new digital industrial economic system of industrial whole industrial chain, whole value chain, whole element wisdom collaboration based open interconnection and scenario integration characterized by virtual promoting real and virtual enhancing real. Industrial Metaverse is one of the typical forms of intelligent manufacturing in the future, which is the advancement and ascension of the digital twin. The current research on industrial Metaverse is in the stage of proof of concept, infrastructure construction, and model exploration, and is in urgent need of a further breakthrough.

(3)  Scenario-driven innovation

Scenario-driven innovation is not only the process of applying existing digital technologies and industrial Internet platform services to specific scenarios to create greater value, but also the process of driving the integration and co-integration of innovation elements and contextual elements such as strategy, technology, platform, and market demand based on future trends and vision needs, breaking through the technical bottlenecks of existing industrial Internet platforms, developing new technologies, new products, new channels, new business models, and even opening up new markets and fields. The scenario-driven innovation provides a breakthrough direction for the industrial digital transformation empowered by the industrial Internet platform. In the future, there should be more focus on the opportunities and mechanisms for carbon peaking, carbon neutrality, common prosperity, military, and civilian integration, and rural revitalization empowered by industrial Internet platforms.

The top three countries in the quantity of core papers in the engineering research front of “research on industrial digital transformation empowered by industrial internet platform” are China, the USA, and the UK (Table 1.2.1). The main output organizations of core papers are Aalto University, South China University of Technology, Beihang University, etc. (Table 1.2.2). From the perspective of the cooperation network among main countries (Figure 1.2.1), there is more cooperation among the USA, China, and other countries. From the perspective of the cooperation network among main institutions (Figure 1.2.2), there is more cooperation among South China University of Technology, King Saud University, and University of Messina.

The quantity of citing papers in China ranks first, as shown in Table 1.2.3. The top-ranked institutions are the Chinese Academy of Sciences, Beijing University of Posts and Telecommunications, and Beihang University, as shown in Table 1.2.4. Figure 1.2.3 shows the roadmap of the engineering research

《Table 1.2.1》

Table 1.2.1 Countries with the greatest output of core papers on “research on industrial digital transformation empowered by industrial internet platform”

《Table 1.2.2》

Table 1.2.2 Institutions with the greatest output of core papers on “research on industrial digital transformation empowered by industrial internet platform”

《Figure 1.2.1》

Figure 1.2.1 Collaboration network among major countries in the engineering research front of “research on industrial digital transformation empowered by industrial internet platform”

《Figure 1.2.2》

Figure 1.2.2 Collaboration network among major institutions in the engineering research front of “research on industrial digital transformation empowered by industrial internet platform”

《Table 1.2.3》

Table 1.2.3 Countries with the greatest output of citing papers on “research on industrial digital transformation empowered by industrial internet platform”

《Table 1.2.4》

Table 1.2.4 Institutions with the greatest output of citing papers on “research on industrial digital transformation empowered by industrial internet platform”

front of “research on industrial digital transformation empowered by industrial internet platform”.

1.2.2 Research on global supply chain security risk management in digital age

The supply chain management refers to business flow, logistics and information flow, including production, transportation, warehousing and many other processes, and any problems in any process can affect supply chain security and cause supply chain risk. In the digital era, digital analysis methods and tools provide the technical guarantee to resist supply chain risks. However, the technology itself is vulnerable to external complex systems, especially under the trend of reverse globalization and active disconnection of supply chains in developed countries. How to ensure supply chain security in global high-tech industries has become an increasingly urgent issue.

Currently, scholars from various countries around the world have provided a quantity of solutions to deal with the global supply chain risks in the digital era. In terms of the main output countries of core papers, the top three countries are the UK, China, and Canada (Table 1.2.5). The main output institutions of core papers are University of British Columbia, University of Guelph, and the Ohio State

《Figure 1.2.3》

Figure 1.2.3 Roadmap of the engineering research front of “research on industrial digital transformation empowered by industrial internet platform”

University (Table 1.2.6), with these three institutions having more than 90 citations.

In terms of the cooperation network among main countries (Figure 1.2.4), there is cooperation between France and Germany, China and Belgium, and the UK and Iran. In terms of the cooperation network among main institutions (Figure 1.2.5), only University of East Anglia and the Ohio State University have not collaborated with other organizations. It can be seen from Table 1.2.7 that the top three countries in terms of the quantity of the citing papers are China, the UK, and the USA, with more than 100 citing core papers each. In particular, the percentage of citing papers in China is as high as more than 20%. Meanwhile, according to Table 1.2.8, the top organizations are Hamad Bin Khalifa University, University of Bordeaux, and Chinese Academy of Sciences.

Based on the review of the existing literature, a more in- depth analysis on how to identify, assess and prevent the risks of global supply chain in the digital era is conducted, and insights are provided on the future development trend. The specific research trends are as follows.

(1)  Risk identification method of global supply chain based on digital technology in complex situations

In the identification of supply chain risks, the more intuitive representation of the digital technology is the big data algorithm, which analyzes the risk characteristics of the supply

《Table 1.2.5》

Table 1.2.5 Countries with the greatest output of core papers on “research on global supply chain security risk management in digital age”

《Table 1.2.6》

Table 1.2.6 Institutions with the greatest output of core papers on “research on global supply chain security risk management in digital age”

《Figure 1.2.4》

Figure 1.2.4 Collaboration network among major countries in the engineering research front of “research on global supply chain security risk management in digital age”

《Figure 1.2.5》

Figure 1.2.5 Collaboration network among major institutions in the engineering research front of “research on global supply chain security risk management in digital age”

《Table 1.2.7》

Table 1.2.7 Countries with the greatest output of citing papers on “research on global supply chain security risk management in digital age”

《Table 1.2.8》

Table 1.2.8 Institutions with the greatest output of citing papers on “research on global supply chain security risk management in digital age”

chain based on the collection of information flow in the supply chain network, establishes a risk prediction system, and predetermines whether there is a security risk in the supply chain. The commonly used supply chain risk identification algorithms include fuzzy comprehensive evaluation method, neural network, etc. The field of application tends to be credit risk, liquidity risk identification and enterprise operation risk prediction in supply chain finance. In the context of the current COVID-19 pandemic, the mainstream of research is to use multiple methods and technologies to identify the risks of manufacturing capacity, production level and other aspects of the supply chain, and to optimize different methods and compare the results.

(2)  Risk impact assessment of global supply chain based on digital technology

Digital technology can be adopted to virtualize and visualize the whole process of the supply chain, and evaluate all potential risks in the supply chain through big data, simulation and other technologies. Recently, the digital technology for assessing the impact of supply chain risks has become more systematic, such as the application of smart devices such as RFID and sensors to collect real-time data. In combination with historical maintenance data and derived data, the above digital technology simulates the operation process of the supply chain in real time and comprehensively evaluates the unknown risks and impacts in the supply chain. The evaluation of supply chain risk through decision support system, tracking system and digital twin technology has become a research hotspot. How to assess the supply chain security risks of global high-tech industries in the current critical period of intensified international trade friction, increased uncertainty and high-quality transformation of various industries has become an increasingly urgent issue under the current trend of reverse globalization and active decoupling of supply chains in developed countries, which deserves further research.

(3)  Interruption and security prevention of global supply chain under digital background

The application of digital technology has led to data co- construction and sharing, promoted multi-chain cooperation, and improved the integrated decision-making ability of the supply chain, which is crucial to risk prevention of supply chain. During the spread of COVID-19, 3D printing, cloud logistics, and contactless distribution technologies and equipment such as automated guided vehicles (AGV) and radio frequency identification (RFID) provide important support for preventing supply chain disruptions. Research hotspots in this field focus on methods to prevent supply chain disruption, and to design and practice supply chain architecture based on blockchain, cloud platform and other technologies. Therefore, the key research directions in academic research recently are layout methods of key technology industrial chain and supply chain, supply chain security problem of high-tech industry under the intensification of reverse globalization economic competition, and supply chain data leakage and information security threat caused by data integration.

(4)  Global supply chain risk impact law and control mechanism driven by data

As an important tool for identifying, evaluating and preventing supply chain risks, the digital technology has been widely studied by scholars for managing supply chain risks, but there remain some aspects to be considered. From the micro level, there are still loopholes in digital technology, with data leakage and information security threats. From the macro level, the impact and the solution mechanism of smart supply chain innovation and government industrial layout on global supply chain risks remain unclear, and it is necessary to prepare a supply chain risk management standard suitable for the digital age. Future research driven by data should focus on solving the problems of weak digital key technologies, specifically, the basic theory and method of supply chain resilience in the digital age, the impact law and Simulation of Sino-US trade friction on global supply chain, the supply chain security assessment and early warning mechanism of key industries in China, the risk of global supply chain disruption, system prediction and security governance system, etc. For the above problems, further study is deserved.

To further improve the global supply chain risk security management in the digital age, this paper further plans the development path of applying digital technology to deal with supply chain risks in the future (Figure 1.2.6).

1.2.3 Research on big data governance methods in artificial intelligence scenarios

By reviewing the existing literature and analyzing the current development situation and future trends, the research on big data governance methods in artificial intelligence scenarios focuses on the following aspects.

(1)   Key technologies for big data governance in artificial intelligence scenarios

The whole process data service includes structured data processing, data quality assessment and data cleaning, data normalization, data storage, and data sharing. Structured data processing starts with parsing multi-source heterogeneous data sources, extracting the required information using information extraction techniques, and further converting them into structured data. The key technology for data quality assessment is data visualization technology. The user defines some data cleaning rules to deal with the quality problems in the data in batches and improve the efficiency of data cleaning. In the field of database research, there are also studies to improve the efficiency of data cleaning with the help of crowdsourcing ideas. In the process of data cleaning, multiple rounds of human-computer interaction are required, and the interaction interface and interaction mode of the system is particularly important for the effectiveness of data cleaning algorithms. The data normalization processing determines the data granularity and expression according to the demand characteristics of the application. It is necessary to consider the balance of big data preservation time and storage space and identify the data elements that have a critical impact on business. It is also desirable to establish a data governance standard system to solve the institutional and technical problems that make data difficult to share.

(2)   Construction of big data mapping knowledge domain in artificial intelligence scenarios

By constructing big data mapping knowledge domain, it realizes understanding data, explaining phenomena and knowledge reasoning, discovering deep relationships, and realizing intelligent search and intelligent interaction. Firstly, based on metadata information such as enterprise and social, algorithms such as natural language processing, machine learning, pattern recognition, and business rule filtering are used to improve processing efficiency and enhance the accuracy of knowledge extraction. Secondly, knowledge is represented and stored in the form of ontology and automatically constructed into mapping knowledge domains.

Finally, through mapping knowledge domain relations, knowledge discovery, knowledge reasoning, and mining are carried out through mapping computing, and intelligent search and correlation query means are used to provide more accurate data to end users.

(3)    Big data security controllability issues in artificial intelligence scenarios

There is a need to solve data security and algorithm security problems such as privacy leakage, data rights validation, algorithm bias, and technology abuse. Firstly, cryptography technology, smart contracts, and privacy computing are the key technologies for achieving data security. Encryption of data using technologies such as public key cryptography can define the identity of data subjects and effectively support data rights validation. Privacy computing technologies, such as multi-party security computing, can realize the exploitation

《Figure 1.2.6》

Figure 1.2.6 Roadmap of the engineering research front of “research on global supply chain security risk management in digital age”

of data without transferring the original data and promote the separation of data ownership and use rights. By integrating privacy computing and trustworthy privacy computing technologies, it can effectively solve the problem of re- identification of personal information after anonymization, and realize the “computable but not identifiable”. Using contract theory and incentive mechanism, it can balance security protection and benefit sharing to realize “data available but not visible”. Secondly, algorithmic models are difficult to explain, difficult to control, and difficult to account for, leading to discrimination, vulnerability, exploitation, information cocoon, and other security risks. By using the four elements including knowledge, data, algorithms, and computing capabilities, it can establish new interpretable and robust AI theories and methods to further develop safe, trustworthy, reliable, and scalable AI technologies.

(4)  Typical artificial intelligence application scenarios of big data governance

Big data governance has been applied in more complex and high-value artificial intelligence scenarios such as finance, healthcare, retail, urban management, public opinion monitoring and so on for solving management needs such as business decision-making, resource allocation, process optimization, operation and maintenance assurance, and risk prevention and control. In the field of business decision- making, big data visualization technology is applied to realize the effective transmission of complex analysis processes and analysis elements. In the field of resource allocation, the dynamic management of resource allocation is realized by relying on big data collection and computing capabilities. In the field of process optimization, bottlenecks in business processes are discovered to improve operational efficiency and customer experience. In the field of operation and maintenance, instant monitoring of operation is realized based on stream data processing technology. In the field of risk prevention and control, risk clues are identified through visualization technology to enhance risk warning capability.

In the above typical AI application scenarios of big data governance, on the one hand, a systematic big data governance framework has not yet been formed, and key technologies such as open sharing, security and privacy protection, quality assessment, and value prediction are far from mature. On the other hand, there are problems including monopoly regulation of Internet companies, regulation of financial digital business, regulation and guidance of online public opinion, data security and privacy protection, and other AI application problems.

The top three countries in the engineering research front of “research on big data governance methods in artificial intelligence scenarios” regarding the quantity of core papers are the USA, the UK, and China (Table 1.2.9), and the top three countries in terms of citations per paper are France, South Korea, and Sweden. Among the main output countries, the USA, the UK, the Netherlands, Australia, and other countries have more collaborations (Figure 1.2.7). The top organizations in terms of the quantity of core papers are Zhejiang University,

《Table 1.2.9》

Table 1.2.9 Countries with the greatest output of core papers on “research on big data governance methods in artificial intelligence scenarios”

Imperial College London, and the London School of Hygiene & Tropical Medicine (Table 1.2.10). Among the main output institutions, Imperial College London and the London School of Hygiene & Tropical Medicine have very close collaborations with other organizations, while University of North Carolina and University of California, San Francisco also have many collaborations (Figure 1.2.8). As can be seen from Table 1.2.11, the USA ranked first in the quantity of citing core papers. As can be seen from Table 1.2.12, the top-ranked institutions are Chinese Academy of Sciences, Harvard University, and the University of Melbourne. Figure 1.2.9 shows the roadmap of the engineering research front of “research on big data governance methods in artificial intelligence scenarios”

1.2.4 Research on the theory and method of accurate construction and evolution of digital twin model

The digital twin is driven by multi-dimensional models and fused data to achieve monitoring, simulation, prediction, optimization, and other services and application requirements, where the construction of the digital twin model is a prerequisite for achieving digital twin practice and landing application. The digital twin model can evolve in real-time by receiving data from physical objects, thus maintaining consistency with physical objects throughout their lifecycle. Based on the digital twin model, analysis, prediction, diagnosis, training and so on can be performed and the simulation results can be fed back to the physical object, thus helping to optimize and

《Figure 1.2.7》

Figure 1.2.7 Collaboration network among major countries in the engineering research front of “research on big data governance

《Table 1.2.10》

Table 1.2.10 Institutions with the greatest output of core papers on “research on big data governance methods in artificial intelligence

《Figure 1.2.8》

Figure 1.2.8 Collaboration network among major institutions in the engineering research front of “research on big data governance methods in artificial intelligence scenarios”

《Table 1.2.11》

Table 1.2.11 Countries with the greatest output of citing papers on “research on big data governance methods in artificial intelligence

《Table 1.2.12》

Table 1.2.12 Institutions with the greatest output of citing papers on “research on big data governance methods in artificial intelligence

《Figure 1.2.9》

Figure 1.2.9 Roadmap of the engineering research front of “research on big data governance methods in artificial intelligence scenarios”

make decisions on the physical object. Models are the core foundation of the digital twin, but modeling technology has emerged since the 1950s and has developed over the decades into more than a dozen new modeling models, technologies, and industries, such as model engineering, data-driven modeling, high-performance modeling, and complex system modeling. The emergence of the digital twin has further promoted the development of modeling technology.

The purpose of digital twin modeling is to eliminate the uncertainty of various physical entities, especially complex systems.

The construction of the digital twin model is the digital modeling of the properties, methods, and behaviors of physical entities and processes in digital space. The model construction can be multidimensional in “geometry-physics-behavior-rules”, or multi-domain in mechanical, electrical, hydraulic, and other fields. The modeling of complex entities is often cross-domain, cross-type, and cross-scale, involving multiple dimensions, and the effect of modeling through a single dimension is not good. It is also necessary to achieve the construction of more complex object models through model assembly and fusion at multiple levels and granularities to achieve a comprehensive characterization of the characteristics of each domain of complex physical objects. To ensure the correctness and validity of the digital twin model, the constructed and assembled or fused models should be validated to check whether the model describes and characterizes the state or features of the physical object correctly. If the model validation results do not meet the requirements, model correction is needed to make the model closer to the actual operation or use of the physical object to ensure the accuracy of the model. Therefore, the main research direction can focus on the six stages of the digital twin model, i.e., “construction-assembly-fusion-validation-correction- management”, for carrying out the accurate construction of the multi-dimensional/multi-domain digital twin model, the assembly and fusion of all-elements/multi-scale twin digital model, the validation and correction of virtual-real consistency of digital twin model, and the interactive iteration and dynamic evolution of digital twin model.

Building digital twin models is not the final purpose, but a means to improve the performance and operational efficiency of their corresponding real objects through the analysis of digital twin models. Important future development directions include multidimensional deep fusion modeling of complex entities, continuous improvement of modeling efficiency and accuracy, interoperability and interactive evolution of models, mutual iteration and dynamic evolution of models and physical entities, and unified model semantics and syntax. The digital twin is in the rising stage of development, the technical system is improving, the industrial integration continues to accelerate, and the industry application is accelerating penetration. Application scenarios include various fields such as aerospace, intelligent manufacturing, health care, smart cities, energy and power, and integrated transportation.

The top three countries for the quantity of core papers in the engineering research front of “research on the theory and method of accurate construction and evolution of digital twin model” are China, the USA, and Germany (Table 1.2.13), and the main output institutions of core papers are Guangdong University of Technology, Beihang University, and the University of Auckland (Table 1.2.14). From the cooperation network of the main countries (Figure 1.2.10), there is very close cooperation between China and other countries, and from the main institutions, Guangdong University of Technology and City University of Hong Kong, National University of Singapore and Beihang University cooperate more closely (Figure 1.2.11). As can be seen from Table 1.2.15, China ranks first in the quantity of citing core papers. As can be seen from Table 1.2.16, the top three institutions are Northwestern Polytechnic University, Beihang University,

《Table 1.2.13》

Table 1.2.13 Countries with the greatest output of core papers on “research on the theory and method of accurate construction and evolution of digital twin model”

and Shanghai Jiao Tong University. Figure 1.2.12 shows the roadmap of the engineering research front of “research on the theory and method of accurate construction and evolution of digital twin model”

《Table 1.2.14》

Table 1.2.14 Institutions with the greatest output of core papers on “research on the theory and method of accurate construction and evolution of digital twin model”