How to foster the innovative development of the platform economy is an important topic of discussion in the current academic community. Digital platforms, as a core component of the platform economy, play a vital role in stabilizing employment, boosting domestic demand, and improving people’s livelihoods. Although existing research has extensively covered areas such as platform monopoly, labor rights protection, and data security, this focus has, unfortunately, overshadowed a holistic understanding of the evolutionary trends, content structures, and logic mechanisms of digital platform governance policies. This has resulted in fragmented and incomplete research findings.
Drawing on the policy tools theory, this study adopts content analysis as the core research method and constructs a three-dimensional analytical framework of "policy structures, governance contents, and policy instruments"to systematically analyze policy texts related to the platform economy at the central government level. A total of 295 valid policy documents were obtained after duplicate screening and relevance evaluation. Three doctoral student members of the research team extracted basic information and core expressions from these documents, refining three analytical units: "content characteristics", "core categories", and "segmented frameworks". This process yielded three major focus themes for digital platform governance, namely platform social obligations, platform social responsibilities, and platform ecosystem, which are further divided into 6 core categories and 18 sub-dimensions. Meanwhile, a supporting segmented framework of policy tools was constructed, covering supply-side, demand-side, and environment-side tools. To ensure coding reliability, four experts conducted independent cross-validation; policy texts with inconsistent coding results were excluded.
This analysis reveals the characteristics of evolutionary phases and governance priorities of China's digital platform governance regulations, addressing the following three questions: What is the evolutionary trend of the policy documents, and what kind of phased changes do they exhibit ?What are the key dimensions and substantive domains of digital platform governance ?What is the logic mechanism of digital platform governance, and what features does it demonstrate?The findings confirm that the evolution of digital platform governance policy has occurred in three stages, with digital platforms being conceptualized successively as 'technical systems' (software platforms), 'bilateral/multilateral markets' (transaction platforms), and 'infrastructure' (platform organizations). In terms of governance focus, the primary emphasis has been on three areas: the platform's social responsibility and social obligation, and the platform ecosystem. In terms of the governing logic, supply-side policy tools, driven by external governance, have compelled digital platforms to fulfill social obligations. Demand-side instruments, centered on internal governance, have incentivized them to take on social responsibility. Environmental instruments, focused on collaborative governance, have encouraged the improvement and optimization of the digital platform ecosystem.
On the basis of the above findings, the study proposes three methods for optimizing policy. First, the government should proactively adjust environmental policy instruments to establish a "system-driven" collaborative governance framework. This can be achieved by innovating governance methods and scaling up proven practices, thereby making the digital platform governance system more coherent and effective. Second, it is essential to optimize supply-side policy instruments and strengthen institutional safeguards to ensure that platforms fulfill their social obligations. This requires innovating policy supply models to incentivize a shift from an "efficiency-first" to a "responsibility-first" paradigm, actively steering platforms towards their social duties. Third, demand-side policy instruments should be innovated to stimulate platforms' internal motivation to undertake social responsibilities. Approaches such as fostering cross-sector integration and launching pilot demonstration projects can help to move beyond a simplistic "punishment-over-regulation" mindset. These measures also promote platform self-regulation and technological innovation, ultimately guiding digital platforms towards "responsible stewardship".

Competition among complementors on digital platforms for limited platform resources and user resources has shown a trend toward involution, profoundly impacting their innovation and performance outcomes. This not only concerns individual survival but also the sustainable development of the platforms themselves. According to resource dependence theory, the competitive pressure generated by resource scarcity has an important impact on organizational behavioral choices and performance achievement. However,studies on complementor performance primarily examine internal attributes, strategic behaviors, and external environmental factors. In contrast, studies on complementor competition tend to focus on the rivalry over user resources, while largely overlooking competition for platform resources. Overall, there is a notable paucity of research investigating how complementor competition within digital platforms affects their performance, as well as the specific mechanisms through which resource competition affects complementor performance, especially from the perspective of complementor experience.
Therefore, drawing on resource dependence theory, the study categorizes resource competition into platform resource competition and user resource competition, and takes apps on the iOS Store APP in China as the research sample to explore the impact of resource competition on complementor performance,further explores the mechanism of iterative agility and iterative novelty, then reveals the moderating role of complementor product experience and ecological experience from the perspective of experiential learning, and finally probes into the influence of resource competition on reputational heterogeneity of complementor performance effects. The study constructs a regression model to verify the proposed hypotheses, and then it adopts a two-way fixed effects model for regression analysis, and incorporates both individual fixed effects and time fixed effects.
It is found that (1)the more intense the competition among complementors for platform and user resources, the greater the degree of inhibition on their performance.(2)The mechanism of action is found to be that this inhibitory effect is realized by reducing the complementors' iterative agility and iterative novelty.(3)Complementor experience serves as a key contingency condition in how resource competition affects performance. Specifically, product experience diversification does not confer an experiential advantage in coping with resource competition; on the contrary, it amplifies the negative impact of such competition. By engaging in multiple concurrent development projects, complementors with diversified product experience suffer from resource fragmentation, insufficient strategic focus, and poor internal coordination. These challenges have weakened their ability to achieve deep vertical specialization and focus on continuous innovation, thereby exacerbating the negative impact on their performance due to the intensified competition for platform and user resources.In contrast, the richness of ecosystem experience enables complementors to have a deeper understanding of platform rules and manage user relationships more effectively. This enables complements to maintain stable market responsiveness and performance even in the context of intensified user resource competition.(4)Increased resource competition has a stronger dampening effect on the performance of complementors with higher reputations.
The main theoretical contributions are as follows: First, the impact of resource competition between complementors entering the platform and other complementors regarding both platform and user dimensions on their performance is explored, which enriches the related literature on the relationship between resource competition and complementor performance. Second, iterative agility and iterative novelty are introduced into the framework of resource competition affecting complementor performance, which elucidates the iterative innovation mechanism by which resource competition affects complementor performance. Third, complementor experience is introduced as a boundary condition, and its moderating effect is explored by dividing it into two dimensions of product experience diversity and ecosystem experience richness in combination with the characteristics of digital platforms, expanding the mechanism of the moderating effect of experience. Fourth, the reputation heterogeneity of the impact of platform resource competition and user resource competition on complementor performance is examined.

With the advancement of the digital economy, the digital transformation of industrial internet platforms under institutional constraints has exhibited structural stagnation. Pharmaceutical commercial platforms, owing to their high dependence on the external institutional environment, are confronted with more complex challenges. Multiple constraints from policies, regulations, industry standards, and social responsibilities have has heightened the complexity of resource integration and business expansion for pharmaceutical platforms while fundamentally transforming their value creation paradigms. Meanwhile, traditional value creation theories, which are rooted in the monopolization of resources and control over single processes, contradict the core nature of platforms (openness, integration, and convergence) and thus fail to guide or explain the evolution of pharmaceutical business platforms. Therefore, it has become an urgent key issue to thoroughly explore how pharmaceutical business platforms achieve value creation through phased open iteration under institutional constraints.
This study employs a single-case research method, using Jiuzhoutong as a case to explore the process of open iteration of platform ecology, the evolution path of service strategies, and the mechanisms of value creation under service-dominant logic. The study adopts a grounded theory methodology, systematically conducting open coding to generate 22 constructs from 31 initial concepts. Axial coding was then performed to identify six core categories, including resource integration, service exchange, institutional arrangements, as well as resource, structural, and rule advantages. Selective coding was subsequently utilized to distill service strategy and value creation as the overarching phenomena. Analytical rigor was enhanced through continuous, "data-logic"cyclic iteration, multiple rounds of member checking with executive participants, cross-source comparative verification, and theoretical saturation testing.
Findings reveal a three-stage open iteration path: bilateral embedding, network expansion, and boundary permeation. Jiuzhoutong's journey demonstrates this path. Its survival-focused bilateral embedding phase targeted underserved primary healthcare/OTC markets, bypassing monopolies via operational innovations: a "Quick Batch-Quick Distribution" model, precision logistics, and cost control, building a foundational physical network. Capitalizing on liberalization, the network expansion phase involved horizontal mergers creating a vast certified warehousing network and vertical integration into TCM cultivation and B2C/O2O e-commerce. Establishing key functional platforms marked its shift to a service provider. Facing tightening regulations and public health demands, the boundary permeation phase featured cross-industry alliances, "Pharma+Tech+Finance" integration, and solutions like its Prescription Outflow Platform. Crucially, this process exhibited dynamic adaptation, forging asynchronous adaptive advantages that sustained growth under constraint and drove service strategy evolution. Service strategy evolution, governed by co-creation imperatives, links iteration to value creation across three phases: service deepening, service radiation, and service coupling. Service deepening established operational excellence and trust through cost control, guaranteed service accessibility, and consistent rules, ensuring stability. Service radiation diffused proven logistics/cost models nationally, adding advanced resource configuration across the pharmaceutical chain and sophisticated data integration via the "FBBC" mechanism, alongside business process standardization and relationship coordination tools, enabling reliable scalability. Boundary permeation necessitated service coupling-integrating diverse service forms beyond provision, manifested in the "Three New and Two Transformations" strategy( which refers to new products, new retail, new medical services, digitalization, and real estate investment trusts) and embedded governance, creating holistic ecosystem solutions. Value creation advanced progressively: transactional value, relational value, and ecological value. Bilateral embedding generated transactional value via scale economies. Building resource advantages, structural advantages, and rule advantages established the transactional base. Network expansion cultivated relational value: enhanced resource advantages from deeper partnerships/data synergy; structural advantages empowering participants; solidified rule advantages as standards became industry benchmarks. Boundary permeation unlocked ecological value-emergent synergy from cross-domain co-creation: resource advantages via cross-industry recombination; structural advantages through distributed, collaborative networks; rule advantages culminating in active industry rule-setting. This delivered significant implicit social value alongside economic returns, including improved pharmaceutical accessibility and reduced health disparities.

As a new strategic competitive unit in the digital economy era, innovation ecosystems evolve through interaction with their environment to achieve sustained innovation.Science-driven innovation relies on extensive interactions among diverse organizations—including core companies, science-based enterprises, universities, venture capital firms, governments, and intermediary institutions—to collaboratively create value, exhibiting highly pronounced innovation ecosystem attributes.Consequently, conducting specialized research on the evolutionary mechanisms of science-driven innovation ecosystems holds theoretical value and practical significance for accelerating the development of emerging and future industries in China and fostering new quality productive forces tailored to local conditions.Current research on innovation ecosystem evolution primarily focuses on the evolution of the system itself, with limited attention to the profound impact of environmental uncertainty.In the VUCA era, the innovation environment undergoes frequent, substantial, and unpredictable rapid changes.Particularly, the rapid advancement of digital technologies renders the external environment of innovation ecosystems dynamic, intense, and complex.Thus, attention cannot be confined to how the system's structure, members, and relationships evolve over time.It is essential to examine the impacts and opportunities that environmental uncertainty brings to the evolutionary process.The core idea is to enable the system to respond to environmental changes rapidly and at low cost, which is termed as the "flexibility of innovation ecosystems".Specifically for science-driven innovation, the high level of environmental uncertainty encountered during the innovation process necessitates adopting "flexibility" as the starting point to analyze the evolutionary patterns of innovation ecosystems.This has given rise to the theme of "flexible evolution of science-driven innovation ecosystems."
　　Grounded in a rigorous definition of innovation ecosystem flexibility and flexible evolution, this study employs a longitudinal exploratory single-case research method grounded in resource orchestration theory.Through re-analyzing the classic case of Huawei, it explores the flexible evolutionary mechanisms of science-driven innovation ecosystems.This study distills a resource orchestration-based model of flexible evolution for science-driven innovation ecosystems.This model describes a continuous transition process from "science-driven innovation components" to "predominantly science-driven innovation" to and then "comprehensive science-driven innovation"which is achieved through the ecosystem's "incubation-optimization-restructuring" evolutionary stages within highly uncertain environments.
The research findings are as follows:(1) The process of flexible evolution is inherently accompanied by a stepwise strengthening of the system's "science-driven" attribute, which goes through three stages:the incubation period of introducing scientific components, the optimization period dominated by scientific drive, and the reconstruction period of comprehensive scientific drive.(2) Flexible evolution is a dynamic adjustment during the process of actively seeking change.Through strategies such as adapting to changes, leveraging changes, and creating changes, it actively adjusts the system, enabling the innovation ecosystem to evolve along the path of chain extension—vertical expansion—all-directional integration.(3) Flexible evolution achieves an innovation capability leap by removing evolution bottlenecks.Through dynamic resource arrangement, it coordinates the relationships among members and continuously removes evolution bottlenecks, allowing the system's innovation capability to dynamically evolve along the path of utilization-driven innovation supporting exploratory innovation—exploratory innovation leading utilization-driven innovation—dual innovation collaborative leap.(4) Flexible evolution obtains conditions conducive to the continuous innovation of the system through dynamic resource arrangement.It is manifested as the architect of the science-driven innovation ecosystem, carefully arranging innovation resources, promoting dynamic resource combination, and obtaining conditions conducive to the continuous innovation of the system in a highly uncertain environment, achieving the flexible effect of seeking benefits and avoiding harms.
This study contributes to theory and practice in three ways:(1) It analyzes the flexible evolution mechanism of science-driven innovation ecosystems, enriching the theoretical framework of science-driven innovation; (2) It advances innovation ecosystem research into a "flexibility"-driven phase, aligning with the digital economy's environmental characteristics; (3) It demonstrates how flexible evolution breaks evolutionary rigidity, driving enterprises to transition from traditional innovation systems to science-driven innovation systems.The research findings offer reference for science-driven enterprise and industrial development, as well as government policy formulation in China.

Against the backdrop of China′s shift toward high-quality development and increasing emphasis on supply chain security, understanding how firms′ information disclosure behaviors influence supply chain resilience has become a critical research issue. While prior literature has primarily focused on technological, financial, or structural determinants of supply chain resilience, less attention has been paid to how distorted innovation disclosures-particularly the misalignment between innovation narratives and actual innovation effort-affect firms′ ability to withstand and recover from disruptions. This study introduces the concept of innovation narrative-action gap into the supply chain resilience literature and examines its impact, mechanisms, and heterogeneity using data from Chinese A-share listed firms from 2012 to 2023.
Drawing on transaction cost theory, the study posits that innovation narrative-action gaps characterize a form of information distortion that undermines inter-firm trust and increases coordination frictions within supply chain networks. When firms exaggerate innovation progress or capabilities in their public disclosures, upstream and downstream partners may form misaligned expectations about the firm′s production capacity, technological reliability, or fulfillment capability. Once these expectations fail to materialize, the resulting trust erosion triggers additional monitoring, renegotiation, and contractual safeguards-raising supply-demand coordination costs and reducing collaborative flexibility. Internally, such misaligned narratives may also distort goal alignment, increase administrative burdens, and generate resource misallocation, thereby weakening firms′ operational agility under disruption. On this basis, the study develops three hypotheses predicting a negative effect of innovation narrative-action gaps on supply chain resilience and two mediating pathways through supply-demand coordination costs and internal coordination costs. To empirically test these predictions, the study constructs a firm-level measure of innovation narrative-action gap using standardized residuals from regressing textual innovation disclosure intensity on future patenting outcomes. Supply chain resilience is measured through an entropy-based composite index that integrates supply chain resistance and recovery capability.
The results reveal that innovation narrative-action gaps significantly undermine firms′ supply chain resilience, and this effect remains robust to alternative variable measurements and a battery of robustness checks, including entropy balancing and instrumental variable approaches. Mechanism analyses indicate that supply-demand coordination costs and internal coordination costs serve as significant mediating channels through which innovation narrative-action gaps impair supply chain resilience. Specifically, firms with larger gaps face greater supply-demand coordination costs as well as higher internal coordination costs due to increased managerial friction and organizational misalignment.
Further heterogeneity analyses show that this negative effect is significantly more pronounced among high-tech firms, and among firms with higher customer concentration and supplier concentration. These findings highlight that innovation narrative-action gaps not only impair internal operational functioning but also propagate across organizational boundaries, weakening the overall resilience of supply chain ecosystems.
The study offers several implications. Firms should enhance the accuracy and verifiability of innovation disclosures and avoid excessive narrative embellishment that may jeopardize long-term collaborative relationships. Policymakers may consider strengthening guidelines and consistency checks for non-financial innovation disclosure to reduce information asymmetry risks. Supply chain partners should incorporate disclosure-execution consistency into their risk assessment frameworks to better anticipate coordination frictions. Overall, this research expands the understanding of how micro-level information distortion behaviors influence supply chain resilience and provides actionable insights for improving both innovation governance and supply chain risk management.

Against the backdrop of the new round of technological revolution and industrial transformation, innovation consortiums led by leading enterprises are the key vehicle for achieving breakthroughs in core technologies and developing new quality productive forces. However, in practice, problems such as difficulty in unifying innovation consensus, obstruction of knowledge collaboration, and poor internal governance are widespread, which restricts the effectiveness of collaborative innovation. The AI for Science (AI4S)-centered artificial intelligence reshapes the research and development model, providing technical support to solve the above problems. Existing studies mostly separately discuss the connotation, establishment mode and innovation performance of the innovation consortium, lacking a complete transmission mechanism that integrates multiple theories to analyze AI's empowerment of multi-party collaboration.
Therefore, this paper combines the scenario-driven innovation theory and the TOE (Technology-Organization-Environment) framework to explain the internal logic of enhancing the efficiency of AI-driven innovation consortiums and proposes practical implementation paths. The research first systematically reviews the core theoretical connotations and applicable scopes of the consortiums, defines the connotation of consortium efficiency from the perspectives of organizational operation, knowledge production, and value transformation, summarizes the three major development dilemmas, and then analyzes the operational scenario-tasked logic. It explains the enabling mechanism of AI in breaking through cognitive, collaborative, and governance bottlenecks from the dimensions of technology, organization, and environment, constructs a "scenario-tasked-TOE three-dimensional interaction-efficiency leap" theoretical model and delineates its progressive transmission relationship. It proposes supporting implementation paths from four dimensions: scenario coordination, intelligent knowledge platform, human-machine agile collaboration, and intelligent governance.
The theoretical deduction yields five conclusions: First, the "unification without integration" of consortiums is motivated by the existence of ambiguous scenario goals, blocked knowledge flows, rigid resource scheduling, and incomplete benefit-risk mechanisms; second, scenarios serve as the starting point of AI empowerment, converting macro strategies into practical research tasks through demand identification, task decomposition, and dynamic resource allocation, and unifying the three-dimensional collaborative goals of the TOE framework; third, AI exerts differentiated empowerment across multiple dimensions: AI4S realizes the digitization of tacit knowledge and breaks through cross-disciplinary integration barriers; human-machine collaboration builds a flat network and dynamically matches innovation elements to reduce coordination costs; and intelligent governance optimizes the distribution of benefits and risk-sharing mechanisms, thereby improving long-term operational systems; fourth, technological, organizational, and environmental dimensions mutually empower one another to form a positive cycle, driving consortiums to achieve triple leaps in structure, cognition, and value, completing the qualitative transformation from simple entity aggregation to deep collaborative innovation; fifth, synchronously promoting the construction of scenario layouts, intelligent knowledge bases, agile collaboration mechanisms, and AI governance systems can fully release the innovation potential of consortiums.
The theoretical contribution of this paper lies in expanding the TOE framework to multi-party, cross-organizational collaborative scenarios, embedding scenario-driven theory into the TOE system to compensate for its shortcomings in static analysis, and building a comprehensive, multi-level theoretical model to clarify the transmission chain from demand to efficacy emergence. Furthermore, the related conclusions can provide practical references for local governments to improve support policies and for chain-leader enterprises to build intelligent, collaborative R&D systems.

Amid a global wave of technological revolution and industrial transformation, the digital-intelligent and green transitions are reshaping the division of labor, operational logic, and competitive paradigms of global industrial chains. Domestically, China grapples with challenges including bottlenecks in the domestic economic cycle, industry-specific overcapacity, weak market expectations, and systemic barriers. However, the rise of the digital-intelligent era has positioned data elements with their non-rivalry, positive externalities, and low-cost replicability as transformative forces, driving enterprises to reduce reliance on traditional factors, optimize data allocation efficiency, and unlock data value to expand production boundaries and enhance core competitiveness. As digital government development advances, the increased openness of government public data has profoundly impacted corporate innovation, inter-enterprise technological linkages, and industrial spatial layouts, creating new opportunities for industrial chain linkages. Yet, empirical research on the value unlocking of data elements remains scarce. Studies on public data openness′ microeconomic effects have largely focused on regional or individual enterprise levels, neglecting the industrial and supply chain dimensions, and the conclusions remain inconsistent.
Addressing this gap, this study exploits the staggered rollout of municipal public data openness platforms across Chinese cities as a quasi-natural experiment and employs a progressive difference-in-differences (DID) approach to analyze data from Chinese A-share listed companies (2008–2023) and examine the impact of public data openness on industrial chain linkage. The dependent variable, industrial chain linkage, is a composite proxy constructed as the product of two dimensions: (1) horizontal linkage, measured at the industry level by production segmentation length which captures the breadth and depth of inter-industry input dependencies and reflects cross-sectoral integration; and (2) vertical linkage, measured at the firm level by specialization degree, which reflects the intensity of upstream-downstream collaboration within the same industrial chain through deepened division of labor. The key explanatory variable is public data openness. In recent years, local government-led open data platforms have served as the primary vehicle for public data disclosure, and their phased implementation provides an ideal setting for empirical identification. Drawing on authoritative reports and academic literature, this study identifies 204 prefecture-level cities that have launched such platforms and manually collects each city′s platform launch year via comprehensive searches using Google, Baidu, and official government websites.
Key findings reveal that public data openness significantly promotes industrial chain linkage, with artificial intelligence applications acting as a critical moderating variable that accelerates this process. Mechanism tests confirm two core transmission channels: cost reduction and expectation stabilization. Heterogeneity analyses reveal that the positive impact of open public data on industrial chain linkages is more pronounced among private enterprises, competitive industries, high-tech industries, and companies in the growth and maturity stages. Further analyses demonstrate that public data openness enhances industrial chain resilience and security, stimulates cross-regional linkage effects, and advances coordinated regional economic development.
This study makes three potential contributions: First, it pioneers the analysis of public data openness as an enabler of industrial chain linkages, leveraging the exogenous shock of public data platform launches to expand relevant literature and offer practical insights for building advanced and efficient modern industrial systems. Second, it clarifies the multi-dimensional pathways (cost reduction, expectation stabilization, risk prevention) through which public data openness influences industrial chain linkages, while exploring heterogeneous effects across contexts, enriching research on industrial chain modernization. Third, by verifying the value-creation effects, resilience-enhancing role, and cross-regional linkage-promoting function of public data openness, the findings provide theoretical guidance for government departments to coordinate public data openness initiatives and for market entities to efficiently utilize public data. They also contribute to unlocking data dividends, advancing the integration of an efficient market and a proactive government, and supporting the security and modernization of industrial chains.

In the context of a new round of scientific and technological revolution,the evolution of industrial technologies has gradually shifted from the breakthrough in a single technology to the coordinated development of multiple interdependent technologies. In complex technological systems,symbiotic technologies are connected through knowledge association,functional complementarity and industrial application. The improvement of one technology may affect the evolutionary direction of related technologies,and further generate a driving effect within the system. However,existing studies on symbiotic technologies mainly focus on formation mechanisms,structural characteristics and co-evolutionary patterns,while insufficient attention has been paid to how driving relationships among symbiotic technologies emerge,how they can be identified,and why their intensity and stability differ.
To address the inadequacies,this paper constructs a value-flow-based analytical framework for identifying symbiotic technology-driven relationships. A multilayer network composed of actors,technology topics and industrial applications is first constructed. In this network,innovation actors provide innovation value to technology topics,while technology topics transfer application value to industrial scenarios. On this basis,the innovation value stream and application value stream of each technology topic are calculated to describe the dynamic value trajectory of technology evolution. Dynamic time warping is then introduced to align the value stream sequences of different technology topics. By comparing the similarity of value flow trends under different time shifts,this paper identifies whether one technology precedes another in the evolutionary process and whether it forms a driving relationship with the associated technology. Furthermore,transfer entropy and the time-offset distribution of DTW alignment paths are used to measure driving intensity and driving stability,respectively. According to these two dimensions,the driving relationships among symbiotic technologies are divided into four types:stable driving,occasional driving,foundational support and temporary coupling.
Taking intelligent chip technology as the empirical field,this paper collects 28 633 authorized invention patents from 2006 to 2025 and uses a dynamic LDA model to identify eight technology topics,including FPGA hardware logic configuration,network device control,SoC chip bus architecture,GPU graphics rendering,processor memory access,video signal processing,target image processing and timing signal control. The empirical results show that a complex driving network exists among symbiotic technologies in the intelligent chip field. Network device control and GPU graphics rendering are the leading technologies,and they drive the evolution of related technologies through multiple paths. Technologies located in key positions of the underlying architecture tend to generate high-intensity and relatively stable driving effects,because their evolution provides basic support for data transmission,module coordination,interface response and computing performance. Foundational supporting technologies may also lead to path dependence,making associated technologies maintain compatibility with the technical standards,functional interfaces and application environments provided by leading technologies. When leading technologies encounter technical bottlenecks or changes in R&D direction,the driving process becomes unstable and shows phased fluctuations or intermittent transmission. In addition,driving relationships mainly realized through indirect technological association are relatively weak in both intensity and stability.
This paper advances the study of symbiotic technology evolution by shifting the focus from static association analysis to dynamic driving relationship analysis. By integrating innovation value streams and application value streams,this paper provides a new perspective for explaining how value is transmitted among symbiotic technologies during technological evolution. Methodologically,the introduction of dynamic time warping makes it possible to capture asynchronous evolutionary patterns and time-lagged relationships that are difficult to identify through conventional correlation methods. In practice,the paper reveals the driving network,leading technologies,driving paths and differentiated driving characteristics in the field of intelligent chips. The findings can help innovation actors identify key technologies,optimize technology portfolios and allocate R&D resources more effectively. They also provide implications for industrial managers to monitor technological bottlenecks,coordinate the development of related technologies and reduce the risk of performance shortfalls in complex product systems.

Under the dual pressures of environmental crises and market competition, enterprises are increasingly compelled to pursue green transformation as an inevitable choice. As a core pathway for achieving this strategic objective, green sustainable innovation is regarded as the key to overcoming the impasse. However, due to practical bottlenecks such as high upfront investment risks and long payback periods, the problem of insufficient internal driving force among enterprises has become increasingly prominent. How to build a long-term incentive mechanism to activate sustained innovation momentum has become a core issue that urgently needs to be addressed in the current process of industrial transformation. Meanwhile, artificial intelligence (AI) technologies, represented by industrial robots, are profoundly reshaping the industrial landscape. Their great potential in improving production efficiency and optimizing resource allocation offers new possibilities for resolving the dilemma of green innovation. Therefore, how does the application of robots affect the level of green sustainable innovation in enterprises? What are the underlying mechanisms? These are the central issues this paper seeks to explore. Although existing studies have examined the relationship between AI and green innovation, in-depth analysis of its impact on sustainable innovation and the specific pathways involved remains insufficient.
This study employs a panel dataset of A-share listed firms in China spanning the period from 2013 to 2023. Firm-level green patent data is matched with industrial robot statistics from the International Federation of Robotics (IFR) to construct measures of robotic application intensity and green sustainable innovation. A two-way fixed effects model is utilized to empirically test the core hypotheses. To further examine the underlying mechanisms, a mediation analysis framework is applied. To address potential endogeneity concerns arising from reverse causality and omitted variable bias, several robustness strategies are implemented, including time trend controls, system GMM estimation, and double machine learning regression.
This study offers several theoretical contributions. First, it broadens the analytical perspective on the green impact of AI by shifting the focus from general green innovation to green sustainable innovation, which holds greater long-term strategic value. Second, a novel dual-pathway analytical framework is proposed to explain how robotic application fosters green persistence, specifically through the upgrading of human capital and the attraction of green investors via signaling effects. Third, the analysis investigates the heterogeneity of this relationship across different contexts, including firm ownership structures, industry characteristics, and the intensity of environmental regulation, thereby enhancing the understanding of the conditional mechanisms involved.
The empirical results indicate that robotic applications significantly enhance firms' levels of green sustainable innovation. Mechanism analyses reveal that this effect operates primarily through two channels: (1) enhancing the level of human capital, and (2) attracting the attention of green investors via signaling effects. Moreover, the positive impact is more pronounced among non-state-owned enterprises, certified green factories, firms in less polluting industries, and those subject to stronger environmental regulatory constraints. Thus, the government should raise the penetration of robots across manufacturing, create dedicated funds and demonstration projects, using tax breaks, subsidies and low-interest loans to incentivize firms to adopt robots for green upgrades. Robots' green-innovation boost is uneven. It is necessary to spotlight green-factory pioneers to set replicable models. For heavy polluters, SOEs firms in weak-regulation regions, where the impact of robot adoption tends to lag,governments should craft targeted programs that pair strict environmental rules with incentives to embed robots in green production, driving full-scale sustainable transformation.
Overall, this study contributes to the growing literature on the green development implications of AI, sheds light on the “black box” through which robotic applications drive long-term green innovation, and offers theoretical grounding and practical insights for firms seeking to leverage intelligent technologies for sustainable development, as well as for policymakers aiming to design effective AI-driven green innovation incentives. Future research could examine whether the synergy between robotics, digital transformation, and AI platforms can amplify green sustainable innovation via energy optimization and process improvements;and researchers could introduce moderating factors such as green finance and leverage quasi-natural experiments or stronger instruments to mitigate endogeneity.

Key core technologies play a decisive role in specific historical periods and industrial domains. Advancing innovation in these technologies is essential for China's transition to a strong science and technology power. However, Chinese latecomer firms face dual constraints in this process: technological blockades from forerunner countries and insufficient indigenous technological accumulation. How these firms overcome the constraints to achieve innovation has become a critical management issue. Existing research has paid insufficient attention to dynamic process mechanisms under different starting points, particularly failing to incorporate technological accumulation into the research context.
Given that theoretical development lags behind China′s cutting-edge innovation management practices, this paper employs a single-case study method. Taking China Railway Construction Heavy Industry Corporation Limited (CRCHI), a leading enterprise in China′s tunnel boring machines industry, as the research case, it explores from a dynamic process perspective: under the dual constraints of technological blockades from forerunner countries and a lack of their own technological accumulation, how do Chinese latecomer firms mobilize complementary innovation entities to collaborate, cobble together multi-source innovation elements, and gradually achieve key core technology innovation? It then aims to distill the dynamic process mechanism of key core technology innovation in latecomer firms.
The study finds that key core technology innovation in latecomer firms is a dynamic process wherein they play a leading role, mobilize complementary innovation entities to collaborate, and thereby create value. Specifically, first, under the guidance of entrepreneurial spirit and guaranteed by innovation institutions, latecomer firms do their utmost to invest in basic innovation elements such as talent and funds to fill the gap of lacking technological accumulation. This serves as the foundational starting point for key core technology innovation and lays the groundwork for subsequently leveraging external forces like the government. Second, in the interactive process of government guidance-firm undertaking, latecomer firms convert external support into motivation to mobilize complementary innovation entities for collaboration. Based on exploratory and search-based innovation element bricolage models, they collaborate with complementary entities to gradually cross two major chasms in key core technology innovation.Third, the progressive connection and cyclical evolution of exploratory and search-based bricolage gradually generate economic, technological, and industrial value. This trifecta marks the successful realization of key core technology innovation by latecomer firms, thereby completing a dynamic closed-loop process.
The findings contribute to theoretical innovation in three aspects: First, by incorporating technological accumulation as a contextual variable, this study identifies the starting point of key core technology innovation in latecomer firms under the dual constraints of technological blockades from forerunner countries and a lack of their own technological accumulation, addressing the theoretical gap in existing research regarding insufficient attention to innovation processes of firms with different starting points. Second, from the perspective of government-enterprise interaction, it reveals the motivational mechanism for latecomer firms to mobilize complementary innovation entities for collaboration, deconstructing the implementation process of latecomer firms how they undertake strategic guidance from the government and mobilize complementary entities for collaboration. This breaks through the limitation of previous studies that primarily focused on the unilateral perspective of government support. Third, centering on the dynamic process of key core technology innovation, the study proposes differentiated innovation element bricolage models for collaboration between latecomer firms and heterogeneous innovation entities, revealing their progressive connection and cyclical evolution. This breaks through the theoretical cognition that treats innovation entities as “homogenized” units. These findings not only deepen the theory of key core technology innovation but also provide theoretical support and practical insights for governments and enterprises to engage in key core technology innovation activities, holding significant importance for China′s strategic goal of achieving self-reliance and strength in science and technology.

Amid stringent environmental regulations and rising survival pressures, firms have developed a heightened awareness of environmental sustainability. However, despite this growing recognition, green initiatives often fail to deliver short-term performance gains because they require high upfront investment, yield slow returns, and involve persistent information asymmetry, thus making it difficult for firms to convert such efforts into measurable outcomes within a limited time frame.Meanwhile, although digital technologies are recognized as catalysts for green innovation and low-carbon transition, they also entail substantial energy consumption and pose a risk of carbon rebound effects. This dual nature implies that pursuing either greening or digitalization alone cannot simultaneously achieve economic performance and environmental governance goals. Against this backdrop, digital-green integration (DGI), which represents the coordinated advancement of digital and green transformation pathways, has emerged as an important strategic route for manufacturing firms.
Existing research on DGI has predominantly emphasized internal governance mechanisms while paying insufficient attention to the contextual influences of external stakeholders such as markets, governments, and the public. Empirical evidence at the micro level regarding the overall impact of DGI on the sustainable development performance (SDP) of manufacturing firms is also limited. Consequently, the mechanisms through which DGI contributes to SDP and the boundary conditions that amplify or constrain its effects remain inadequately understood. These research gaps call for a more comprehensive analytical framework that integrates internal mechanisms with external stakeholder pressures.
To address these issues, this study draws on strategic synergy theory and stakeholder theory to examine how DGI affects the SDP of manufacturing firms and how market-oriented, governmental, and societal stakeholders shape this relationship. Specifically, the study incorporates multiple forms of external stakeholder influence including supply chain pressure, competitive pressure, government regulatory pressure, R&D subsidy incentives, and public attention into the analysis. Using panel data from Shanghai and Shenzhen A-share listed manufacturing firms between 2009 and 2023,the study employs two-way fixed effects models and moderation analyses to investigate the effects of DGI on financial and ESG performance and to evaluate the moderating roles of these external pressures.
The findings demonstrate three key insights. First, DGI improves both financial performance and ESG performance, thereby enhancing overall SDP, and these results remain consistent across multiple robustness tests. Second, different external stakeholders play heterogeneous moderating roles. Competitive pressure and public attention strengthen the positive effect of DGI on SDP, suggesting that firms respond more proactively to DGI initiatives when competitive intensity and public visibility are high. In contrast, R&D subsidy incentives weaken this relationship, indicating that subsidies may induce opportunistic behavior or reduce firms’ incentives to invest in substantive capability building. Meanwhile, supply chain pressure and government regulation reinforce the impact of DGI on ESG performance but do not significantly alter its relationship with financial performance. Third, heterogeneity analysis reveals that the effects of DGI vary markedly across managerial expertise, ownership type, industry pollution intensity, and levels of environmental uncertainty. The positive effect of DGI on financial performance is stronger in firms whose senior managers possess digital or environmental expertise, in state-owned enterprises, and in firms operating under low environmental uncertainty. For ESG performance, the positive effect is more pronounced in non-state-owned enterprises and in firms facing lower environmental uncertainty. Additionally, DGI enhances financial performance in non-polluting industries but dampens it in heavily polluting ones, likely due to higher compliance costs and stricter environmental constraints in these sectors.
The implications emerge from these findings. Governments should refine institutional frameworks, strengthen the supervision of subsidy implementation, and establish more effective regulatory mechanisms to encourage firms to advance DGI in a substantive and efficient manner. Manufacturing firms should develop adaptive DGI strategies that align with external stakeholder expectations and their internal resource endowments, transform external pressures such as supply chain demands, competitive dynamics, regulatory oversight, and public scrutiny into sustainable competitive advantages, strengthen managerial capabilities in ESG stewardship and digital transformation, and tailor DGI pathways to their organizational characteristics and evolving market conditions.

In the context of rapid digital transformation, enhancing employees' digital creativity has become a vital driver for organizations to overcome the "digital dilemma" and achieve sustainable innovation. Despite escalating global investments in digital transformation, a significant value-deployment gap persists, wherein employee distrust and passive utilization of artificial intelligence (AI) constitute critical bottlenecks impeding technology-enabled value creation. Extant research has predominantly centered on the instrumental attributes of AI or its negative emotional repercussions, while relatively neglecting employees' subjective perceptions regarding AI's impact on their professional futures, namely the dual-dimensional structure of AI awareness (challenge vs. hindrance), and its underlying mechanism in influencing digital creativity through autonomous motivation. Meanwhile, the ethical dilemmas inherent in AI applications serve as pivotal contextual factors, yet their moderating effects remain insufficiently examined.
Accordingly, this study adopts a dual theoretical framework, integrating the Job Demands-Resources (JD-R) Theory and Self-Determination Theory (SDT), to explore the dual impact of AI awareness on employees' digital creativity. Specifically, a moderated mediation model is constructed to examine the mediating role of autonomy motivation and the moderating effect of moral dilemmas in shaping these relationships. This study employs a multi-wave survey methodology, drawing upon 257 valid responses from employees in technology-driven enterprises across various regions in China, to investigate the influence mechanism of AI awareness on digital creativity. The sampling strategy balances regional diversity with technological advancement, encompassing representative cities across different geographical areas and economic development tiers, including Beijing, Shanghai, Chengdu, and Chongqing. Data collection was conducted in two stages: At Time 1 (December 2024 to January 2025), the study measured AI awareness, autonomy motivation, and moral dilemmas; At Time 2 (March to April 2025), the study assessed digital creativity, thereby effectively mitigating common method bias and ensuring rigorous causal inference.
The findings reveal the following key results: (1) AI challenge appraisal significantly and positively influences autonomy motivation, which subsequently enhances digital creativity. Conversely, AI hindrance appraisal exerts a significant negative influence on autonomy motivation, suppressing digital creativity. (2) Moral dilemmas negatively moderate the relationship between AI challenge appraisal and autonomy motivation, reducing the positive effect of AI challenge appraisal on digital creativity through autonomy motivation. However, moral dilemmas do not significantly moderate the relationship between AI hindrance appraisal and autonomy motivation. (3) The moderating effect of moral dilemmas underscores the ethical complexities employees face when interacting with AI, as such dilemmas introduce additional cognitive and emotional barriers to leveraging the technology.
The findings offer actionable insights for organizations striving to optimize their use of AI. To harness the potential of AI as a driver of innovation, organizations must focus on fostering positive perceptions of AI among employees. This includes highlighting the ability of AI to improve efficiency and innovation while addressing employees' ethical concerns and apprehensions. Managers can mitigate the negative effects of AI hindrance appraisals by implementing regular training programs, workshops, and case-sharing sessions that emphasize the benefits of AI and provide hands-on experiences to reduce fear and anxiety. Simultaneously, organizations should establish ethical guidelines and provide psychological support to employees facing moral dilemmas related to AI applications, such as concerns over data privacy or algorithmic fairness. Moreover, creating a supportive work environment that emphasizes autonomy is essential for sustaining innovation.
In conclusion, this study sheds light on the dualistic impact of AI awareness on employees′ digital creativity, emphasizing the importance of balancing the enabling and constraining effects of AI in the workplace. By fostering autonomy, reducing hindrance perceptions, and addressing ethical concerns, organizations can unlock the full potential of their workforce in achieving digital innovation. Future research should extend this work by examining these dynamics across different cultural contexts and over longer time horizons to gain deeper insights into the evolving relationship between AI and workplace creativity.

Technology mergers and acquisitions (M&A) are widely used by firms to acquire external knowledge and enhance innovation. However, the failure rate of technology M&A remains high, at approximately 70% in both the U.S. and China. Existing academic research on technology M&A innovation mostly focuses on the integration of technical talent, organizational structure, and resource allocation within acquired firms, or explores the optimization of M&A decision-making from the perspective of resource complementarity and technical matching between the two parties, yet largely overlooks the vital impact of M&A activities on the key inventors of the acquiring firms themselves. These key inventors serve as the core carriers of the acquiring firm′s original technical accumulation and the main force for knowledge absorption, integration, and secondary innovation following an M&A; changes in their creativity during the post-M&A stage directly determine the efficiency of benefit transformation and the sustainability of corporate innovation.
This study takes the evolution of creativity among key inventors in acquiring firms after technology M&A as its core research object, constructing a dynamic multi-level creativity evaluation model to realize scientific and accurate assessment of their creativity evolution, thereby addressing the limitations of existing single-dimensional and static evaluation methods. By combing the theories of social network, innovation economics and talent creativity, this study builds a multi-dimensional and interrelated evaluation framework from three core levels: collaboration network dynamics, individual inventor characteristics, and M&A context features, which systematically covers the internal and external factors affecting post-M&A creativity.
In terms of research methods, this study adopts a two-stage PCA-BP neural network model to solve the problems of high dimensionality and multicollinearity of initial indicators, as well as the non-linear correlation between influencing factors and creativity changes: firstly, Principal Component Analysis (PCA) is used to reduce the dimension of 12 initial evaluation indicators, filter out redundant information and extract core components; secondly, the extracted principal components are input into the Back-Propagation (BP) neural network for training and fitting, so as to build an efficient and accurate non-linear evaluation model.
For empirical testing, this study selects 219 key inventors active between 1997 and 2024 as the research sample, collecting 21 616 valid patent records with key inventors identified by patent quantity, citation frequency and claim counts to ensure sample representativeness. The sample is divided into 170 training sets and 49 testing sets to verify the model′s effectiveness. The test results show that the PCA-BP model has a mean square error (MSE) of only 0.00058 and a prediction accuracy of 97%, with its predicted creativity trajectory highly consistent with the actual change trend, demonstrating strong generalization ability and robustness. Comparative analysis further proves that the dynamic multi-level index system and PCA-BP model are significantly superior to traditional static indicators, single BP neural network and random forest model in evaluation accuracy, convergence speed and anti-interference ability.
This study realizes innovation in research perspective and method in the field of technology M&A innovation: in theory, it shifts the research focus from acquired firms to the core inventors of acquiring firms, putting forward a systematic “network dynamics-individual characteristics-M&A context” analysis framework and enriching the theoretical research on post-M&A talent creativity and innovation management; methodologically, it integrates PCA and BP neural network, providing a feasible technical path for non-linear evaluation in small-sample and high-dimensional scenarios. In practice, the constructed model can serve as a targeted diagnostic tool for enterprises to monitor the creativity changes of key inventors after M&A, helping enterprises timely identify creativity decline risks, carry out targeted intervention, optimize core human capital management, and further improve the innovation performance and success rate of technology M&A, thus providing important theoretical support and practical reference for M&A decision-making and post-merger integration management.

The deep integration of digital technologies into lean management has become a strategic priority for manufacturing firms. However, the coexistence of ambiguous terms,such as lean digitalization, Lean 4.0, and lean automation,has led to conceptual fragmentation and blurred boundaries. Current research on Lean Digitalization remains in a preliminary stage, characterized by conceptual ambiguity and fragmented frameworks. The coexistence of inconsistent terminology, such as "Lean 4.0" and "Lean Automation," without unified definitions leads to construct proliferation and limits the comparability of studies. This lack of clarity often results in practical misunderstandings, where digital tools are merely superimposed on existing processes rather than being integrated with core Lean principles. Structurally, existing frameworks are often either too theoretical to be operational or too narrow to be systemic, failing to explain the underlying mechanisms between dimensions. Consequently, these gaps hinder empirical validation and prevent the provision of a comprehensive implementation roadmap for enterprises. Thus, this study aims to conceptually redefine lean digitalization in the manufacturing context with clear and reusable boundaries, and empirically identify its structural dimensions and interrelationships, thereby providing a foundational framework for subsequent measurement development and empirical testing.
This study adopts a two-phase sequential design. In Phase 1, a semantic decomposition method was applied to systematically compare and evaluate existing definitions from 11 representative studies, following established rules for conceptual clarity. In Phase 2, a grounded theory approach was employed, using semi-structured interviews with 31 practitioners from diverse manufacturing sectors, job levels, and ownership types. Secondary data (e.g., company documents, public reports) were used for triangulation. Data coding followed open, axial, and selective coding procedures, with theoretical saturation verified on reserved transcripts.
The analysis reveals three major shortcomings in existing definitions: insufficient term specification, conflation of definition with performance outcomes, and conceptual ambiguity. In response, this study proposes a refined definition: Lean digitalization in manufacturing enterprises is the process of systematically reshaping lean principles, tools, and implementation activities by integrating information, computing, communication, and connectivity technologies to improve operational systems and value creation activities. This definition removes contextual dependency on specific industrial paradigms (e.g., Industry 4.0), avoids teleological presuppositions of success, and retains the core lean logic.From the grounded analysis, six structural dimensions emerge: (1) strategic planning for lean digitalization, (2) IT infrastructure, (3) end-to-end collaboration, (4) intelligent flow optimization, (5) intelligent abnormality control, and (6) human-centered management. These dimensions are systematically integrated into a "Lean Digitalization House" framework: strategic planning forms the roof; IT infrastructure and human-centered management constitute the dual foundation (technical and organizational); intelligent flow optimization and intelligent abnormality control serve as the two core operational pillars; and end-to-end collaboration extends internal capabilities to the value chain. Each dimension is further delineated into sub-categories, such as predictive maintenance, smart built-in quality, lean digital talent development, and supplier synchronization.
This study makes three theoretical contributions. First, it provides a conceptually robust definition of lean digitalization that is decoupled from transient Industry 4.0 rhetoric and performance presuppositions, thereby enhancing construct clarity and cross-study comparability. Second, unlike prior fragmented frameworks, it empirically validates a holistic, multi-dimensional structure that explicitly distinguishes strategic, infrastructural, operational, and collaborative layers. Third, by grounding the framework in diverse manufacturing contexts, it reveals how digital technologies serve not as standalone tools but as enablers that amplify lean principles rather than replace them. From a practical standpoint, the framework offers a stratified roadmap for manufacturing firms: top management can use the strategic planning dimension for evaluation and goal setting; middle managers can leverage IT infrastructure and end-to-end collaboration for capability building and cross-boundary integration; and shop-floor teams can apply intelligent flow optimization, intelligent abnormality control, and human-centered management for daily improvements. The findings also highlight contextual contingencies, suggesting that small and medium-sized enterprises should start with minimum viable infrastructure, whereas large firms need stronger data governance and standardization.