The remainder of this paper is structured as follows: section “Theoretical framework and evolutionary model” details the Cellular Automata model design; section “Simulation analysis” presents simulation outcomes and analysis; and section “Discussion and conclusion” discusses theoretical contributions and policy implications for optimizing AI cluster development.
The theoretical framework of this study integrates two complementary theories: evolutionary economic geography theory and complex systems theory, with a particular focus on Cellular Automata (CA) modeling and simulation. This integrated framework seeks to offer a comprehensive understanding of the dynamics within clustered innovation ecosystems, while also establishing the theoretical foundations for collective learning and competitive advantage in the context of AI-driven manufacturing innovation.
Evolutionary economic geography is a theoretical approach that views the economic landscape as a complex adaptive system. It highlights the roles of competition, spatial environment, technological change, and historical time in shaping economic regions through processes such as variety, selection, retention, and the spatial dynamics of firms and industries. This theory synthesizes Darwinian principles with additional concepts, such as plasticity, robustness, and self-organization (Martin and Sunley, 2006).In this theory, the spatial concentration of related economic activities is seen as a key driver of innovation. The clustering of firms and industries facilitates innovation through mechanisms such as knowledge spillovers, collaborative learning, and the sharing of resources, all of which contribute to a regional competitive advantage. Within the context of AI-driven manufacturing innovations, this theory suggests that the resources within the cluster, inter-firm networks, and the regulatory environment of the cluster collectively influence the emergence and diffusion of innovations. Specifically, the proximity of firms within a cluster enables knowledge exchange and collaborative learning, while the regulatory environment plays a critical role in shaping or limiting incentive structures and knowledge-sharing practices.
In this study, we define manufacturing industry clusters that utilize AI technologies as geographically concentrated networks of enterprises, institutions, and resources that integrate artificial intelligence (AI) to foster growth and innovation within the manufacturing sector. These clusters consist of advanced manufacturing enterprises, research institutions, funding mechanisms, and collaborative networks, all dedicated to applying AI solutions aimed at optimizing production processes, streamlining supply chains, and driving product innovation(Lee et al., 2018; Rizvi et al., 2021; Zdravković et al., 2021). The defining characteristics of these AI-enabled manufacturing clusters include: the integration of AI technologies, such as predictive maintenance, smart manufacturing, automation, and data analytics, to enhance operational efficiency and productivity. Access to shared resources, which include specialized knowledge, skilled labor, digital infrastructure, and financial capital tailored to AI applications. Collaborative networks that facilitate knowledge exchange, technology adoption, and co-innovation among manufacturing firms, AI developers, and research institutions. Supportive environments, characterized by favorable policies, market dynamics, and socioeconomic factors, that encourage the adoption and scaling of AI technologies within manufacturing clusters (Haricha et al., 2023; Jagatheesaperumal et al., 2022; Wan et al., 2020).
Drawing on Evolutionary Economic Geography Theory and synthesizing existing research, this paper presents a comprehensive model outlining the primary drivers of artificial intelligence (AI) innovation and its subsequent evolution within manufacturing industry clusters. Figure 1 provides a visual representation of this model. As depicted in Fig. 1, the progress of AI innovation in manufacturing clusters is underpinned by three primary driving forces.
Cluster resources, encompassing human capital, financial capital, digital infrastructure, R&D capabilities, and corporate culture, are vital for fostering AI innovation within manufacturing enterprises. This resource composition, including human capital, financial assets, and digital infrastructure, is visualized in Fig. 2, which illustrates the key components driving cluster innovation potential. Strategic Human Resource Planning (HRP) enhances the value of clusters by focusing on core talent, thereby boosting competitiveness (Shen, 2011). Capital resources impact industrial clusters by shaping market competition and boosting the vitality of financial enterprises through the development of financial industrial clusters (Ali et al., 2014). Digital resources contribute to industrial clusters by integrating companies into innovative cluster structures, which in turn create favorable environments for introducing digital economy elements in urban settings (Ivanenko et al., 2019). The construction of infrastructure positively influences industrial clusters by attracting firms, facilitating industrial transformation, enhancing regional innovation, driving economic growth, and supporting the transition to industrial centers. The effectiveness of this infrastructure depends on local conditions, such as the availability of skilled labor and capital (Wu et al., 2023). R&D capabilities are key to attracting new firms and enhancing both innovation and performance within industrial clusters (Leung and Sharma, 2021). Additionally, corporate culture plays a pivotal role by fostering employee engagement, driving innovation, and enhancing competitiveness. It stimulates entrepreneurship, shapes policy development, and facilitates the introduction and diffusion of technology (Putilova and Shutaleva, 2020).
In the context of AI-driven manufacturing innovations, the abundance of resources within clusters positively influences the exploration of new innovations by firms operating within the cluster innovation ecosystem. According to evolutionary economic geography theory, the spatial concentration of related economic activities promotes innovation through mechanisms such as knowledge spillovers, collaborative learning, and the sharing of resources. As a result, firms located in resource-rich clusters are more inclined to pursue new AI-driven manufacturing innovations, thanks to the availability of specialized knowledge, skilled labor, and advanced technological infrastructure. It is evident that manufacturing industry clusters play a critical role in fostering the development of AI innovation.
The industrial cluster network represents a complex, localized system of interconnected enterprises, characterized by free-scale properties. This network fosters innovation, resource integration, and economic development through mechanisms such as corporate risk appetite, collaborative knowledge sharing, and strategic cooperation (Abhari et al., 2019). As depicted in Fig. 3, these components constitute the core framework of the cluster network. Within this context, the structure of the industrial cluster network significantly influences the efficiency of knowledge diffusion and innovation (Schilling and Phelps, 2007). Additionally, enterprise risk preferences shape behaviors and strategic choices within industrial clusters, influencing levels of cooperation and necessitating risk prevention strategies for sustainable development (Liu and Xu, 2018). Knowledge sharing within these clusters enhances economic performance, innovation, and competitiveness by reducing costs and risks, and is influenced by factors such as commitment and leadership (Wang et al., 2017). Furthermore, strategic cooperation within industrial clusters promotes firm clustering, supports industry formation, enhances regional innovation, and increases competitiveness. These strategies vary depending on industry type and regional conditions (Baldassarre et al., 2019).
While geographic proximity offers advantages for resource pooling and interaction, it does not, by itself, guarantee the global competitiveness of firms within clusters (Lee, 2018). The ability to build and leverage international linkages is a critical factor for firms seeking access to global markets, advanced technologies, and diversified talent pools (Rugman et al., 2012). Moreover, the composition of clusters, characterized by similar and complementary firms, significantly influences their innovation dynamics. Competitive firms often drive each other toward greater efficiency and innovation, while complementary firms foster collaborative advancements by integrating expertise and resources across the value chain (Hermundsdottir and Aspelund, 2021; Maciel and Fischer, 2020). These interactions underscore the essential role of well-structured inter-firm networks in cultivating an environment where knowledge sharing serves as a catalyst for sustained innovation.
The strength of inter-firm networks positively influences the dissemination of AI-driven manufacturing innovations among firms within the clustered innovation ecosystem. Robust and adventurous inter-firm networks facilitate knowledge exchange, collaborative learning, and joint problem-solving elements critical to the diffusion and adoption of AI-driven manufacturing innovations. Firms embedded in well-connected networks can leverage these relationships to access valuable information and resources, thereby enhancing their capacity to propagate innovation (Aviv et al., 2019). Consequently, it is imperative to strengthen the establishment of cluster networks to promote the advancement of manufacturing industrial clusters and facilitate their transition to a more sophisticated stage.
The industrial cluster environment encompasses a variety of factors, including the economy, policy, industry standards and specifications, and market demand, all of which positively influence talent growth within the clusters (Weng, 2008). These factors within the cluster environment are graphically depicted in Fig. 4. The economic environment plays a crucial role in the development of industrial clusters by fostering talent growth, contributing to economic expansion, enhancing the business environment, and integrating clusters into global networks (Narayana, 2014). State-supported policies can promote the growth and development of industrial clusters, leading to economic benefits and poverty alleviation (Shakib, 2020). Industry standards enhance the development of industrial clusters by improving competitive advantage, overcoming bottlenecks, and influencing economic development, innovation, and digital transformation (Li and Wu, 2016). Market demand affects the development of industrial clusters by lowering average costs and impacting regional economic growth, which can subsequently lead to increases in GDP, labor value, land and property prices, and environmental consequences (Wang, 2018).
A supportive cluster environment positively influences collective learning among firms within the cluster’s innovation ecosystem, thereby contributing to regional competitive advantage. Such an environment shapes incentive structures and fosters a knowledge-sharing atmosphere that facilitates collaborative learning and innovation among firms. Policies and institutions that encourage cooperation, protect intellectual property rights, and provide funding for research and development can motivate firms to engage in collective learning activities, ultimately enhancing the cluster’s competitive advantage (Tallman et al., 2004). The effective utilization of resources and networking within manufacturing industry clusters relies on a favorable political and economic environment, which acts as a catalyst for the integration of artificial intelligence (AI) into the innovative evolution process within these clusters.
Complex systems encompass a variety of phenomena in nature, characterized by interactions among multiple factors (Corning, 1995).In the realm of economics, complex systems theory emphasizes analysis from a connectivity perspective, concentrating on value creation through new connections among elements (Foster, 2005).
Complexity theory, an emerging field since the 1980s, builds upon traditional theories such as control theory, information theory, and systems theory, while also integrating newer theories including dissipative structure theory, synergetics, disaster theory, chaos theory, and super cycle theory. Society itself is a complex system, and social simulation is increasingly recognized as a vital method for understanding the dynamics of social development in the age of artificial intelligence. At the heart of social simulation lies the construction of ’emergence’ mechanisms, which accurately depict and illustrate the processes and outcomes of social operations. The term ’emergence’ refers to properties or capabilities of a system that are not inherent in its individual components (Hegselmann and Flache, 1998).
By utilizing emergent mechanisms to interpret complex social systems, social simulation elucidates intricate relationships across various systems and layers within society, as well as the resultant social processes. In the 1990s, sociologist Luhmann employed the concept of ‘system complexity’ to analyze complex social behavior, demonstrating that strong correlations among elements in a social system can lead to phenomena such as self-organization and self-generation. Social systems exemplify typical complex systems, and their evolutionary patterns are neither entirely fixed nor completely random; rather, they cluster within specific boundaries and ’emerge’ when surpassing a critical threshold. Consequently, emergence is a fundamental characteristic of complex social systems. Social simulation can be integrated with big data and extensive datasets to transcend mere mechanical data aggregation and analysis, thereby revealing the complex properties and evolutionary patterns inherent in these systems.
In this study, we employ Cellular Automata (CA) as a modeling tool to simulate the dynamic and emergent behaviors within AI-enabled manufacturing clusters. First introduced by Von Neumann and Ulam in the 1950s (Crutchfield, 2011). CA has been widely used across fields such as economics, sociology, and ecology due to its ability to capture the interactions between simple components that give rise to complex system behaviors. For instance, in the context of Industry 4.0, Cellular Automata-based models have been utilized to optimize Big Data processing while minimizing energy consumption, as evidenced by the cost-effective MapReduce model proposed by Mitra (2021). This work simplified the complexity of existing CA rules, facilitating efficient data shuffling and integration within industrial processes (MITRA, 2021). In both computational research and industrial applications, Cellular Automata have been effectively employed to exploit their inherent parallelism for high-performance computing on modern platforms, such as multiprocessors and GPUs, yielding significant modeling outcomes (Wa̧s & Sirakoulis, 2015; Was and Sirakoulis, 2015).
A manufacturing industry cluster represents a complex social network, characterized by the intricacies of its evolutionary processes and the challenges posed by the lack of dynamic data for quantitative analysis. This makes it particularly suited for studying innovation ecosystems like manufacturing clusters, where localized interactions between firms can lead to significant collective outcomes. By utilizing CA, we can model how AI technologies diffuse through a cluster and predict the evolution of innovation within these ecosystems. However, the capacity of Cellular Automata (CA) to simulate these evolutionary processes through image descriptions, combined with the powerful functions of MATLAB, renders it a suitable tool for this analysis.
The Cellular Automata simulation model developed in this study is specifically tailored to capture the unique dynamics of AI-enabled manufacturing clusters. These clusters are defined as geographically concentrated networks in which AI technologies are deeply integrated into manufacturing enterprises, research institutions, and collaborative ecosystems. The model incorporates three primary dimensions- resources, networks, and environments- into a computational framework that simulates the adoption and evolution of AI-driven innovations.
Cluster resources are represented by parameters such as human capital, AI infrastructure, and R&D capacity, which collectively influence the integration of AI technologies and firm-level innovation. Collaborative networks are modeled as interactions among manufacturing firms, represented by neighboring cells on the Cellular Automata grid, which emulate AI-driven knowledge sharing, strategic alliances, and competitive dynamics. Lastly, cluster environments are captured through external factors such as government policies, market demand, and economic conditions, all of which shape the adoption of AI technologies and influence the collective behavior of firms within the cluster.
This Cellular Automata framework is particularly well-suited for modeling AI-enabled manufacturing clusters, as it captures emergent behaviors driven by localized interactions. For instance, firms equipped with advanced AI tools and significant resource endowments can stimulate innovation in neighboring firms through spillover effects. Furthermore, supportive policies and a favorable economic environment create conditions that encourage collective learning and risk-taking across the cluster. By simulating the interplay between these dimensions, the model offers a detailed understanding of how AI technologies drive innovation, optimize resources, and transform traditional manufacturing clusters into hubs of technological advancement.
By aligning the model parameters and rules with the defining characteristics of AI-enabled manufacturing clusters, this study provides both theoretical and practical contributions. The results yield actionable insights for policymakers and industry leaders, highlighting strategies to enhance cluster resources, strengthen inter-firm networks, and create supportive environments for AI integration. This approach not only advances theoretical understanding of AI-enabled manufacturing ecosystems but also serves as a guide for fostering industrial competitiveness and innovation in the age of artificial intelligence. The findings underscore the transformative potential of AI technologies in driving systemic innovation and regional economic development.
The Cellular Automata (CA) model developed in this study is structured as follows:
The probability of state transition is governed by several parameters:
For a full breakdown of model parameters and their distributions (e.g., μ, r, and e), refer to Table 1, which summarizes the parameter settings aligned with theoretical foundations and prior literature.
Complex systems theory, grounded in Cellular Automata modeling and simulation, provides a robust analytical framework for capturing the intricate dynamics and nonlinear interactions among the components of the innovation ecosystem(Chopard et al., 2002). By simulating the behaviors and adaptation patterns of complex systems, such as innovation clusters, this theory helps predict the emergence and diffusion of AI-driven manufacturing innovations. Cellular Automata modeling enables the examination of the collective behavior of firms within a cluster, taking into account the complex interplay between cluster resources, inter-firm networks, and the cluster environment.
By synthesizing the principles of evolutionary economic geography theory with complex systems theory, informed by Cellular Automata modeling, the proposed theoretical framework elucidates the intricate interplay between cluster resources, inter-firm networks, and the cluster environment, as well as the emergence and diffusion of AI-driven manufacturing innovations within industry clusters. The unit of analysis is the firm, which is embedded within the broader innovation ecosystem and engages in collaborative learning and competitive interactions with other firms. The theory encompasses a feedback loop of exploration and exploitation, wherein firms utilize cluster resources and inter-firm networks to pursue new innovations. Simultaneously, the cluster environment shapes the development of incentive structures and knowledge-sharing mechanisms that either facilitate or hinder such exploration.
The proposed theoretical framework builds upon and extends existing theories in evolutionary economic geography by incorporating insights from complex systems theory and Cellular Automata modeling. This integration allows for a more nuanced understanding of the dynamics within clustered innovation ecosystems, capturing the nonlinear interactions and emergent properties of the system. Consequently, the framework offers a comprehensive explanation of how cluster resources, inter-firm networks, and the cluster environment collectively influence the emergence and diffusion of AI-driven manufacturing innovations.
In conclusion, this theoretical framework serves as a valuable lens for examining the dynamics of clustered innovation ecosystems and the theoretical foundations of collective learning and competitive advantage in the context of AI-driven manufacturing innovation. By integrating evolutionary economic geography theory with complex systems theory, informed by Cellular Automata modeling, the framework enhances our understanding of the intricate interplay among cluster resources, inter-firm networks, and the cluster environment in shaping the emergence and diffusion of innovations. This framework can also guide future empirical research and inform policy formulation aimed at fostering innovation. AI-driven innovation clusters play a crucial role in enhancing regional competitiveness. The proposed agent-based model integrates these elements to simulate the dynamics of clustered innovation ecosystems, as well as the theoretical foundations of collective learning and competitive advantage in the context of AI-driven manufacturing innovation. This model investigates the complex interplay between cluster resources, inter-firm networks, and the cluster environment, alongside the exploration-exploitation feedback loop that propels the emergence and diffusion of AI-driven manufacturing innovations within industry clusters.
