The Spatial Effect of Industrial Intelligence on High-Quality Green Development of Industry under Environmental Regulations and Low Carbon Intensity

Abstract

In order to thoroughly investigate how industrial intelligence influences green industrial development through direct, indirect, and spatial spillover effects in China and fill in the gaps left by earlier studies, the study combines industrial intelligence and green industrial development into a single analytical framework. The findings show that implementing industrial intelligence can proactively encourage high-quality green industrial development; additionally, a strong spatial correlation is shown between industrial intelligence and high-quality green industrial development. According to spatial spillover analysis, industrial intelligence fosters the development of green industries both inside and between regions. When regional heterogeneity is analyzed, it is revealed that the eastern part of China experiences industrial intelligence effects more strongly than the central region, while the western areas are unaffected. Environmental regulations are a crucial mediating mechanism for the operation of industrial intelligence; in particular, public-participation environmental regulation and market base environmental regulations strengthen the baseline relationship; however, industrial intelligence does not impact high-quality green industrial development through administrative environmental regulation. The partial mediating effect of carbon intensity was also observed. The findings could be used as a guide for decision-making by experts and policymakers in China and other developing nations to use industrial intelligence and support the green development of the sector during economic transformation.

1. Introduction

Recent years have seen a tremendous increase in China’s economic growth, making China an essential representative of the newly industrialized countries [1]. The standard of living and overall well-being of the population has significantly increased. However, such quick development has been accompanied by several unexpected social issues, including an imbalance in the industrial infrastructure, environmental pollution, and the need to maximize energy efficiency. According to statistics from the Global Environmental Performance Index 2020, China is ranked 120th and 130th regarding ecosystem vitality and the Global Environmental Performance Index [2]. In light of the global perspective, China must urgently consider the seriousness of its environmental and ecological difficulties to build a green, ecological society. The main difficulty facing China as it enters a new developmental phase is the growth of high-quality green industries.

Environmental regulation is the pivotal point in achieving green and intensive industrial growth. Because the environment and resources are considered public goods, employing market processes to solve environmental problems has some restrictions. Therefore, environmental regulations must be put into effect by government departments. Moreover, as part of sustainable green development, lowering carbon intensity is another important objective, which has gained substantial political importance worldwide. Due to the severity of climate change, low-carbon development is becoming increasingly popular, and innovative technologies are urgently needed to promote economic growth that emits carbon emissions [3]. Many businesses are incorporating environmental preservation and sustainable development into their strategic decisions due to the growing market demand for low-carbon products [4]. Su et al. [5] and Umer et al. [6] believe that environmentally friendly technologies protect natural resources, produce energy effectively, and reduce carbon intensity. In order to achieve energy conservation and emission reduction, Motorola has built a sound environmental management system and invested money and human resources in the stages of raw material selection, production, and product packaging (The case study is at http://www.31dt.cn/dtqy/2010-03-31/171.html 20 October 2022). Numerous nations have also launched several programs for reducing carbon emissions (CER), such as China’s Green Manufacturing Engineering Implementation Guide (2016–2020) and the European Emissions Trading System, both publishing rules and regulations restricting carbon emissions (Xu et al., 2016). According to the “Survey Report on Energy Conservation and Emission Reduction of Chinese Enterprises Since the 13th Five-Year Plan”, 10% of Chinese businesses are unable to complete the nation’s mandatory annual emission-reduction plan. Moreover, Hao et al. [7] proved a strong correlation between low carbon intensity and green development.

An approach to sustainable development known as “green development” focuses on protecting the natural environment in connection to resource-carrying capacity and the ecological environment [8]. However, the emphasis on green development in industrial operations is green industrial development [9]. The industrial revolution and the new wave of technology are driven mainly by intelligent technology, created from breakthrough technology, which works as a new production tool. Industrial intelligence is an approach to development where industries use intelligent technology to provide output increments and a mode of human–computer interaction. The United States and Germany have consecutively introduced “industrial Internet” and “Industry 4.0” initiatives to change the industrial structure using industrial intelligence. In order to fully capitalize on this historic opportunity, China, the world’s largest developing nation, has introduced a strategy to build the advanced phase of artificial intelligence to promote the close connection between the real economy and intelligent technology. In light of the background, the recent extensive pollution and the fastest industrial revolution raise a number of natural and intriguing queries. Does industrial intelligence contribute to China’s high-quality green industrial development? If industrial intelligence impacts high-quality green industrial development, is there regional difference exist? Do environmental regulations and carbon intensity impact the relationship between industrial intelligence and high-quality green development of the industry, and to what extent? The main objective for this paper is to answer these questions.

According to previous studies, industrial intelligence strengthens the relationship between the research and application sectors, which is a new engine for advancing green technologies [10]. Regarding this, many academics and organizations have thoroughly investigated how industrial intelligence affects economic growth [11] and labor structure [12]. Sun et al. [13] categorized labor into three types: skilled, entrepreneurial, and educated. The authors noted that industrial intelligence benefits entrepreneurial and educated workers more than skilled workers. Acemoglu and Restrepo [14] indicated that using robots would negatively impact the labor force’s employment and decrease the overall number of jobs in manufacturing in Germany. Further, industrial intelligence encourages the replacement of labor with senior and junior school diplomas with advanced machinery, which causes the “polarization” of the Chinese labor market [15].

However, the current academic study of industrial intelligence is concentrated on how it affects the labor structure, economic benefits, and production effectiveness. Whether and how industrial intelligence impacts green industrial growth is still up for debate. Our findings contribute to the literature as follows: (1) the research deepens and broadens the current literature by further refining the idea of industrial intelligence employing many aspects, including intelligence of logistics and services, production processes, production factors, and warehousing; (2) three views are being used to build the system of green industrial development: social, ecological, and economic benefits; (3) whether industrial intelligence impacts the industry’s high-quality green development is investigated and tested using different econometric models; (4) this study gives an analysis with a focus on spatial spillover effects between regions and heterogeneity; (5) from the standpoint of Environmental regulation and carbon intensity, the study investigates how industrial intelligence affects the green development of the industry through transmission mechanisms.

2. Development of Hypothesis

Impact of Industrial Intelligence on High-Quality Green Industrial Development

Digitalized information and knowledge are production variables that directly enhance the viability of urban green development [16]. According to studies, the basic digital technologies used in industries can drastically lower transaction costs and aid businesses in gaining a competitive edge [17]. Enterprises merely want to increase production efficiency through intelligent production, despite the fact that their limited intelligence makes it difficult to achieve scale and efficiency increases in the short term and that they ignore green institutional innovation and management change, both of which are harmful to improvements in green efficiency [18]. In addition, industrial intelligence offers a productive information exchange platform for industry development, removes supply and demand imbalances, increases the market scale and access to information channels, realizes resource integration and sharing among diverse industries, firms, and regions, and systematizes resource allocation. The new generation of artificial intelligence is progressively developing into an emerging dividend promoting green growth, in contrast to conventional dividends, such as land and energy, which have a diminishing marginal influence [19], reducing pollution and resource utilization with effective operating efficiency. In order to encourage the growth of a sustainable economy, digital technology may direct the creation of green products and technical innovation that will increase production efficiency and alter consumption patterns [19]. On the basis of artificial intelligence technology, large-scale expansion in industrial development mood will shift to intensive growth, with human–machine cooperation serving as the primary production strategy and knowledge and data serving as the first production components [20]. The case for developing a sustainable green industry can be further strengthened by focusing on how intelligent technologies are advancing new sectors. Based on the analysis above, we constructed our first hypothesis:

Hypothesis 1.

Industrial intelligence positively impacts high-quality green industrial development.

A new vitality has been infused into China’s economic model with the rapid development of the digital economy based on digital technologies [21]. Information networks and data resources, which establish an open ecology and overcome geographical barriers, are important carriers and production elements of the growth of the advanced economy in comparison to conventional business practices [22,23]. Intelligent technologies are being applied to industry in ways that promote cross-regional information flow and interaction and go beyond traditional economic boundaries, which might have a spatial spillover impact. The rapid expansion of digital and information-based technologies into many industries also hastens the creation of new strategies for resource utilization that support excellence in development [18,24]. Patterns of regional economic growth have altered dramatically as a result of the depth of social integration and penetration of intelligent technologies, making the digital economy a huge opportunity for nations to increase economic competitiveness [25]. In addition, transregional labor and cooperative agreements can be directed by intelligent technology [26], increasing market fairness and transparency and forcing businesses to pursue green innovation for long-term sustainable success in the growing market. According to Wielgos et al., big data resources can be used to lessen the information imbalance between the demand and supply sides, advancing the development of the digital innovation system of information exchange to support green products [27]. Ramaswamy and Ozcan [28] argued that intelligent technologies could successfully link consumers and manufacturers to increase output and resource efficiency. Focusing on previous studies, we assume that the degree of industrial intelligence in one area will likely have an impact on green industrial development in other regions. Therefore, we hypothesize that:

Hypothesis 2.

High-quality green industrial development in neighboring regions may be impacted spatially by industrial intelligence.

Chinese provinces differ regarding technological advantages, resource endowment, and economic development [29], evidencing regional heterogeneity. Dan and Qing argued that the use of digital technologies has eliminated past barriers to economic activity and encourages communication and exchanges between cities and regions beyond distance and time [30]. Therefore, regional heterogeneity might have an impact on the stimulating influence of industrial intelligence on green industrial development. Additionally, Meng et al. [31] claim that the western region exhibits a less prominent substitution effect and value-added benefit of industrial intelligence than the middle and eastern regions. The likelihood of acquiring digital dividends is higher in developed regions than in emerging areas due to their competitive advantages in developing and implementing digital infrastructure and their excellent technological advantages [32]. Moreover, regional economies’ capacity to grow and sustain themselves is impacted by variations in advanced innovation capabilities and technical advantages [33]. The study proposed the below hypothesis in view of the above theories:

Hypothesis 3.

Industrial intelligence has a heterogenous spillover effect on high-quality green development of inter-regional industries.

The industry’s adoption of new intelligent technologies helped the Chinese economy to switch from traditional economic growth to innovation-driven forms, opening up new development areas [34]. According to Ribeiro-Navarrete et al. [35], digital transformation may encourage companies to test out new business models or progress their production technology. It may also compel polluting companies to advance environmental protection and green development [36]. OECD research (2009) indicates that environmental policies significantly contribute to the advancement of green development [37]. In the 1980s, the Chinese government began enforcing environmental laws or regulations as environmental issues grew intensified. These regulatory measures have put much pressure on China’s sustainable industrial green growth, especially in areas with severe resource shortages and environmental deterioration. According to Liu et al. (2016), environmental regulation supports technical advancement resulting in a situation where the ecological environment, the green economy, and green development all gain. Using the American manufacturing sector as an additional example, Carrion-Flores [13] concluded that there was a clear inverse relationship between environmental technology patents and pollutant emissions, and that environmental regulation implementation in the US was helpful in promoting green technology in businesses. Therefore, we believe that the environmental regulation pressure may indirectly result in better industrial green growth performance. Correspondingly, we propose:

Hypothesis 4.

Industrial intelligence promotes high-quality green industrial development through environmental regulations.

A new industry is sped up, and production efficiency is increased as a result of the invasion of intelligent technology into actual industries. Rapid innovation and industrial growth have necessarily led to increased resource use, carbon intensity, and other negative effects, which harm the green development of the industry. Wiebe and Yamano [38] argued that demand-based emissions are required to achieve green growth. In order to speed up the cohesive development of the economy and environment, local governments and environmental protection businesses get funds from national agencies for the R&D of green technology [39]. Moreover, to foster a regulatory framework that supports green growth, local governments will gradually release essential rules and regulations on carbon emission reduction and green technology [40]. Two general categories can be used to describe research findings about how intelligent technologies affects carbon intensity. One side argues that developing industrial intelligence can lower carbon intensity emissions, while the other side claims that these developments help to increase carbon emissions. The majority of researchers favor the first kind of conclusion; Meng et al. [32], for example, indicated that industrial intelligence inhibited the level of carbon intensity. Moreover, Xu et al. [41] pointed out that energy efficiency improvements help in reducing carbon intensity.

Similarly, Yu and Du [42] highlighted how a reduction in CO₂ emissions is directly related to the development of energy technology innovation. Additionally, environmental technology advancements in energy production can support a green economy [4,5]. Considering the abovementioned theories, we infer that industrial intelligence can help achieve more effective energy production, resource conservation, and low carbon intensity, stimulating economic growth without increasing environmental pollution. Therefore, we hypothesize that:

Hypothesis 5.

Carbon intensity indirectly impacts the connection between industrial intelligence and high-quality green industrial development.

Based on the analysis above, Figure 1 presents a research framework for examining the connection between dependent, independent, and mediating variables.

3. Methodology

3.1. Variables

3.1.1. High-Quality Green Industrial Development

Green development includes growing a green economy and upgrading industrial structures, enhancing social welfare, and increasing economic efficiency [43]. The study employed three major indicators: social benefits, ecological and environmental benefits, and industrial green economic benefits, with eleven measurement indicators to assess regional industrial green development systematically. Appendix A provides a list of the specific content. The social benefits of an enterprise focus on making people feel satisfied and content with the progress being made together. The most obvious signs of a person’s quality of life are employment, income, and consumption. The use of the social security fund, the urban registered unemployment rate, the per-capita consumer expenditure of residents, and the disposable per-capita income of residents are all used as indicators of the social benefits of green industrial development.

The industry’s ecological and environmental benefits largely include improved ecological environments and environmental construction. Environmental construction is reflected by environmental pollution control efficiency (EPCE) and environmental pollution control intensity (EPCI). EPCE was evaluated using the overall solid waste utilization rate, and the proportion of investment used to control industrial pollution served as a proxy for environmental pollution control intensity EPCI. Next, to measure the ecological environment’s green development, the study used two indicators: (1) pollution emission efficiency and (2) energy consumption efficiency. Solid waste, carbon dioxide, and wastewater emissions of 10,000 Yuan GDP are used to quantify the efficiency of industrial pollution emissions. Energy consumption efficiency is evaluated using the elasticity of electricity consumption and elasticity of energy consumption.

Industrial economic benefits are typically measured by the international market competition effect, the impact of science and technology innovation, and the spatial agglomeration effect [44]. International competition’s effect is reflected by the industry’s capacity for competition on the global market. The value of regional industrial imports and exports is divided by the value of regional output to calculate it. The impact of science and technology innovation focuses largely on the configuration factor brought about by business model innovation, technological innovation, and management during industrial development. It is determined using both the total number of patent applications and the ratio of R&D investment to GDP.

The locational entropy method quantifies the spatial agglomeration effect [45], which depicts how industrial resources are coordinated in regional distribution. The following is the calculating formula: $\mathrm{ISi}=\frac{\mathrm{Si}/\mathrm{Pi}}{\mathrm{Ei}/\mathrm{E}}$.

Province’s secondary employment is denoted by si, the total employment of the province is symbolized by Pi, Ei stands for national secondary employment, and total national employment is denoted by E. A region’s high technology industry concentration is higher with the location entropy index.

3.1.2. Industrial Intelligence

Industrial intelligence is a growing field that strongly emphasizes the legitimate application of intelligent technologies, such as the Internet of Things, big data, and artificial intelligence, into industrial production processes. The industrial intelligence evaluation method incorporates the popularity of intelligent software, intelligent equipment investment, innovation capability, and information collecting capability [46]. The industrial intelligence system was evaluated in the current study using data processing, and software maintenance capabilities considering intelligent transformation requires the effective use of digital resources. The principal component analysis was used to measure each province’s relative level of industrial intelligence. The detailed content is shown in Appendix B.

3.1.3. Environmental Regulation

Researchers have assessed environmental regulation strength using either the comprehensive or type indices. The present studies assume that the majority of current measurement techniques begin from the standpoint of a single environmental control investment or use a composite index of the number of “three wastes” (solid waste, wastewater, waste gas) treatment projects, failing to develop an index from the function of various tool kinds. Therefore, in accordance with the research conducted by Ren et al. [47], this article employs the heterogeneous environmental regulations index, and we classified environmental regulations into three categories: (1) public participation-environmental regulation, (2) market-based environmental regulation, and (3) administrative environment regulation.

Public-Participation Environmental Regulation (PER) is the process in which the general public participates in environmental regulation operations by communicating their environmental interests. Due to the limitations imposed by government environmental regulation legislation, China’s avenues for public participation in environmental regulations are limited. The only ways are either through environmental letter visits or by filing ecological complaints. Due to the high degree of randomization and unpredictable nature of environmental prosecution data, the number of letters per citizen about the civil environmental petition is used to measure public-participation environmental regulation.

China’s primary market-based environmental regulation (MER) strategies are sewage subsidies, sewage levies, and tradeable licenses. In light of China’s stable and long-established sewage policy, we measured market-based environmental regulation using the ratio of income from sewage fees to the total value of industrial output. The term “administrative environmental regulation” (AER) refers to environmental policies, laws, and regulations adopted by environmental protection organizations or government agencies. The most common and mandatory environmental regulation in China is administrative environmental regulation. It can be split into three categories based on the various regulations’ stages: pre-incident, incident, and ex post [48]. The first pre-incident regulation in China is the “three simultaneous” method, which allows us to measure administrative environmental regulation using the ratio of the “three simultaneous” projects’ environmental investment to the total value of industrial production.

3.1.4. Carbon Intensity

Controlling industrial carbon emissions is essential for China to reach carbon neutrality targets due to the country’s high degree of industrialization and the fact that industrial activities produce 70% of its carbon emissions. Therefore, the study used CO₂ emission per unit of GDP to quantify carbon intensity. The industrial use of fuel oil, kerosene, coke, natural gas, crude, diesel, oil, and gasoline served as the basis for our computation of carbon intensity using the carbon emission coefficient approach [49].

3.1.5. Control Variable

The following choices are made for control variables to thoroughly evaluate the industrial intelligence impact on green high-quality industrial development: (1) Urbanization: this measurement is based on the logarithm of population density; (2) Economic development: the region’s per capita GDP logarithm indicates economic development [50]; (3) Government intervention: the study divides the fiscal expenditure with GDP to measure Government intervention; (4) Financial development: the ratio of financial institution deposits and loans to GDP is used to assess financial development; (5) Openness: the study converted real applied foreign capital into RMBs using the year’s average RMB/USD exchange rate [51], and the percentage of foreign capital used to GDP is used for specific measurement.

3.2. Research Model

3.2.1. Direct Impact Model

Constructing a benchmark regression model to examine industrial intelligence’s direct influence on the industry’s high-quality green development is the first step in validating the research hypothesis

lnGhd_it = α₀ + α₁lnII_it +α₂ Control_it + μ_i + θ_t + ε_it

(1)

where Ghdi represents a high-quality green industrial development, II denotes industrial intelligence. i is the year, and t symbolizes province. Control denotes the control variables, including Urbanization, Economic Development, Government intervention, Financial development, and Openness. Time and regional effect are represented by μ_i and θ_t, respectively, and ε_it indicates the random error.

3.2.2. Model of Spatial Econometrics

To analyze how industrial intelligence impacts high-quality green industrial development across various regions, a spatial econometric model was built. In accordance with the relevant objects, researchers have created various spatial econometric models. However, the study used the spatial Durbin model (SDM) to capture both exogenous and endogenous spatial interaction effects because the spatial error model (SEM) and spatial autoregressive model (SAR) only account for one type of spatial interaction impact. The following SDM model is developed to thoroughly study the mechanism by which industrial intelligence affects high-quality green industrial development:

lnGhd_it = λ₀ + ρWlnGhd_it + λ₁W lnII_it + λ₂lnII_it + λ₃WControl_it + λ₄Control_it + μ_i + δ_t + ε_it

(2)

where WlnGhd_it represents the high-quality green industrial development as having a spatial lag, WlnII_it symbolizes the spatial lag of industrial intelligence; the control variables’ spatial lag term is denoted by WControl_it, and p indicates the spatial spillover effect of the dependent variable in the study.

3.2.3. Mediating Effect Model

To determine if environmental regulation and carbon intensity mediate the connection between industrial intelligence and high-quality green industrial development, the following mediating effect test models were created. The PROCESS plug-in of SPSS.24 is used in this study to examine the mediating effect.

lnGhdi_i = β₀ + β₁lnII_it +β_k Control_it + μ_i + θ_t + ε_it

(3)

lnMed_it = γ₀ + γ₁lnII_it +γ_k Control_it + μ_i + θ_t + ε_it

(4)

lnGhdi_i = φ₀ + φ₁lnII_it + φ₂lnMed_it +φ_k Control_it + μ_i + θ_t + ε_it

(5)

The mediating effect is tested following Wen’s three-step method (2004). At the initial stage, β₁ is tested; if β₁ is significant, the following steps are carried out. In the second step, γ₁ and φ₂ are tested, and a significant mediating effect is considered if both γ₁ and φ₂ are significant. In the third step, to ensure the partial or full mediating effect, φ₁ needs to be examined. The mediating variable exerts its full mediating influence if φ₁ fails the significance test, implying that industrial intelligence affects the high-quality green industrial development through mediating variable. However, if φ₁ meets the criteria for significance, the mediating variable has a partial mediating effect illustrating that both direct and indirect channels are involved in how the intelligent development of industries influences high-quality green development of the industry. If neither γ₁ nor φ₂ is significant, the significance of the mediating effect is assessed using the bootstrap sampling method.

3.3. Data

This study has chosen data from 31 Chinese provinces, covering the years 2010 to 2020. The China Electronic Information Industry Statistical Yearbook, the Population and Employment Statistical Yearbook of China, and the Science and Technology Statistical Yearbook of China were the sources of the initial data.

Table 1. Descriptive statistics.

Variables	Symbols	Interpretation	N	Mean	SD	Minimum	Maximum
Explained variable	Ghd	high-quality Green development of industry	2025	0.533	0.427	−0.417	1.850
Explanatory variable	II	Industrial Intelligence	2025	0.857	0.612	1.011	2.5990
Mediating variables	ER	Environmental Regulations	2025	0.614	0.345	0.132	2.140
	CI	Carbon intensity	2025	0.420	0.527	0.012	2.106
	Urb	Urbanization	2025	4.997	1.137	1.553	7.901
	ED	Economic development	2025	9.661	0.542	6.993	12.542
Control variables	GI	Government intervention	2025	2.109	1.824	1.147	15.001
	FD	Financial development	2025	2.813	1.990	0.619	31.622
	Ope	Openness	2025	0.516	0.718	0.1001	26.229

portrays the descriptive statistic for each of the variables analyzed in the study. The findings illustrate that the average Ghd value from 2010 to 2020 is 0.533, and the maximum and minimum values are 1.850 and −0.417, respectively, demonstrating considerable differences across different provinces. Additionally, the standard deviation of industrial intelligence (II), the environmental regulations (ER), carbon intensity (CI), and the control variables indicate all differ significantly. However, each variable’s average value falls mainly within the normal range, reflecting a suitable sample size and pointing to a reliable conclusion.

4. Results and Discussions

4.1. Direct Effect Test

The baseline regression results are computed using model (A), and the findings are presented in

Table 2. Results of Baseline Regression.

Variables	lnGhd	lnGhd	lnGhd	lnGhd	lnGhd	lnGhd
lnII	0.4421 *** (0.0533)	0.1371 *** 0.0144)	0.1547 *** (0.0272)	0.1338 ** (0.0227)	0.0421 *** (0.0470)	0.0212 *** (0.0368)
lnUrb		0.0531 *** (0.0439)	0.0262 *** (0.0556)	0.0701 * (0.0173)	0.0336 ** (0.0535)	0.0672 *** (0.0981)
lnED			0.0237 *** (0.0529)	−0.2610 (−0.3421)	−0.5711 ** (−0.2290)	−0.5457 *** (0.3671)
lnGI				−0.7201 ** (0.0347)	−0.5512 *** (−0.0912)	−0.7551 * (−0.3141)
lnFD					−0.0336 (−0.0021)	0.0725 *** (0.0044)
lnOpe						0.6635 ** (0.0561)
Constant	1.2271 ***	0.5443 ***	−2.0112 *	4.3443 ***	3.3391	4.5644 **
R2	0.2331	0.6549	0.7710	0.2344	0.1090	0.2231
observations	2025	2025	2025	2025	2025	2025

*, **, and *** show the significance level at 10%, 5%, and 1%; t-values are shown in brackets.

. Industrial intelligence efficiently drives high-quality green industrial development, according to the calculated coefficient of industrial intelligence, which is consistently positive regardless of the contribution of the set of control variables. It may be because intelligent technologies make the industry more environmentally friendly and sustainable by lowering operating expenses, increasing operating effectiveness, and switching out high-energy production elements with low-energy ones. Results verify Hypothesis 1. Our results are consistent with earlier research, which concluded that intelligent technologies help enterprises in the upgrading of management equipment and pollution monitoring efficiently [1]. Furthermore, industrial intelligence stands out for its complementing and substitutive effects on low-skilled workers. In other words, industrial intelligence aids the organization, management, and oversight of resource allocation and production activities, which makes it easier to increase energy efficiency and productivity [52].

4.2. Spatial Empirical Results Analysis

4.2.1. Spatial Correlation Test

The spatial correlation between high-quality green industrial development and industrial intelligence was examined using the global Moran’s index. The spatial results are estimated using the adjacency matrix (W1) and the geographic matrix (W2) in order to minimize estimation error brought on by incorrectly set spatial weights. According to findings in

Table 3. Moran index of industrial intelligence and high-quality Green development of industry.

Year	II				Ghd
	W1		W2		W1		W2
	MI	Z values	MI	Z values	MI	Z values	M I	Z values
2010	0.153 ***	10.442	0.013 **	5.339	0.022 **	3.343	0.231 **	0.762
2011	0.197 **	9.017	0.044 ***	5.761	0.034 ***	3.351	0.242 ***	0.339
2012	0.227 ***	10.361	0.024 ***	6.092	0.071 **	3.440	0.206 ***	0.251
2013	0.176 ***	10.220	0.028 **	5.772	0.101 ***	4.360	0.198 *	1.362
2014	0.244 ***	11.392	0.031 ***	5.322	0.021 ***	4.324	0.265 ***	1.356
2015	0.253 ***	10.001	0.029 ***	5.656	0.043 ***	3.291	0.190 ***	0.212
2016	0.141 ***	9.910	0.034 ***	6.721	0.051 ***	4.268	0.203 ***	1.462
2017	0.210 **	11.326	0.041 *	7.390	0.121 *	4.341	0.218 **	1.247
2018	0.137 ***	12.431	0.037 ***	7.251	0.055 ***	3.411	0.229 ***	1.391
2019	0.210 ***	10.732	0.022 ***	6.109	0.024 ***	2.406	0.241 ***	0.971
2020	0.264 ***	11.441	0.029 ***	5.365	0.041 ***	3.387	0.188 ***	0.543

*, **, and *** show the significance test at 10%, 5%, and 1%.

, under various weight matrices, the significant positive coefficients of Moran’s index (MI) suggest that the spatial econometric analysis may be further investigated.

4.2.2. Spatial Econometric Model Regression

The spatial effect estimation findings are presented in

Table 4. Spatial effect estimation results.

Variables	lnGhd SDM (W1)	lnGhd SDM (W2)	lnGhd SAR (W1)	lnGhd SAR (W2)
lnII	0.1341 ** (0.0241)	0.1421 *** (0.0341)	0.2522 *** (0.02325)	0.2370 ** (0.0355)
W*lnII	0.6332 ** (0.3911)	0.4722 *** (0.1991)
DE	0.1341 ** (0.0441)	0.1421 *** (0.0357)	0.2415 *** (0.0447)	0.2413 ** (0.0565)
IE	0.2215 ** (0.4140)	0.4211 ** (0.6122)	−0.0141 ** (0.0341)	−0.0161 *** (0.0512)
TE	0.3556 ** (0.5103)	0.5632 *** (0.6575)	0.2274 * (0.4671)	0.2252 ** (0.5219)
Control variables	yes	yes	yes	yes
observation	2025	2025	2025	2025
Log-pseudolikelihood	101.2478	198.3317	317.0016	315.4207
R2	0.5093	0.4412	0.7132	0.8682

*, **, and *** show the significance test at 10%, 5%, and 1%; t-values are shown in brackets.

in light of various spatial weight matrices. The SDM model’s spatial interaction terms and all of the industrial intelligence coefficients are positive, confirming the spatial spillover effect of industrial intelligence on high-quality green industrial development. Next, the study divided industrial intelligence’s impact on high-quality green industrial development into direct and indirect effects to better comprehend the spatial spillover effect [53], considering that the point estimate test of the spatial spillover effect could create model estimation bias. The total effect (TE), indirect effect (IE), and direct effect (DE) are all significantly positive, as shown in Table 4, implying that industrial intelligence may efficiently boost green growth in a region while also advancing high-quality green development in neighboring areas via spatial spillover effects. Results confirm Hypothesis 2.

4.2.3. Heterogeneity Analysis

In order to examine the regional variations, we separated the sample into western (Shaanxi, Guizhou, Sichuan, Yunnan, Chongqing, Qinghai, Ningxia, Gansu, and Xinjiang), eastern (Shanghai, Beijing, Hebei, Zhejiang, Jiangsu, Tianjin, Liaoning, Fujian, Hainan, Guangxi, Guangdong, and Shandong) and central (Jilin, Inner Mongolia, Anhui, Jiangxi, Shanxi, Heilongjiang, Henan, Hunan, and Hubei) regions. This was conducted in light of the enormous differences in resource endowment, economic development, and policy implementation of Chinese provinces.

Table 5. Regional regression effect.

Variables	Western	Central	Eastern
lnII	0.0405 (1.523)	0.0646 *** (3.0217)	0.1477 *** (5.1632)
Direct effect	0.0437(1.6630)	0.0651 *** (4.872)	0.1481 *** (6.2203)
Indirect effect	0.0221 * (2.1045)	0.0391 *** (4.3451)	0.1127 *** (4.1263)
Total effect	0.0658 * (3.4413)	0.1042 *** (5.2281)	0.2608 *** (6.1705)
Control variables	yes	yes	yes
R2	0.8521	0.9371	0.8931
observation	2025	2025	2025

*, **, and *** show the significance test at 10%, 5%, and 1%; t-values are shown in brackets.

indicates that industrial intelligence substantially impacts the high-quality green industrial development of central and eastern regions but has no impact on the western region. Results confirmed Hypothesis 3 that the industrial intelligence’s impact on high-quality green industrial development varies by region, supporting Meng et al. [31] research conclusions. Explanation illustrates that the popularity of intelligent software, availability of relatively high-tech expertise, ample chances of Intelligent equipment investment, and advanced innovation resources left the western region behind central and eastern regions. Western regions have a higher share of industrial units that consume a lot of energy, emit emissions, and produce high levels of pollution, but technology is underdeveloped, which impedes the development of green industries.

4.3. Mediating Effect

The findings of the SDM model and baseline regression show that industrial intelligence helps improve the industry’s green high-quality development in China. This section will examine whether environmental regulation and carbon intensity play an intermediary role in the relationship between industrial intelligence and high-quality green industrial development.

Table 6. Mediating impact of environmental regulations.

Variables	(1) lnGhd	Med = MER		Med = AER		Med = PER
		(2) lnMER	(3) lnGhd	(4) lnAER	(5) lnGhd	(6) lnPER	(7) lnGhd
lnII	0.6247 *** (6.2610)	0.2019 *** (2.3135)	0.5521 (4.1091)	0.3130 ** (2.3317)	0.5211 ** (4.601)	0.2720 *** (3.2910)	0.5742 ** (5.3013)
lnMER			0.0431 ** (1.1053)
lnAER					0.0552 (2.7102)
lnPER							0.0463 ** (2.2261)
constant	4.2201 ***	2.3443 **	5.1091 ***	3.6121 ***	5.001 **	3.9121 **	6.2218 ***
Control v	yes	yes	yes	yes	yes	yes	yes
R2	0.5231	0.533	0.6101	0.4281	0.592	0.591	0.6103
Fixed effect	yes	yes	yes	yes	yes	yes	yes
observation	2025	2025	2025	2025	2025	2025	2025

*, **, and *** show the significance test at 10%, 5%, and 1%; t-values are shown in brackets.

shows the estimation outcomes for the transmission mechanism. Column (1) indicates the significant positive effect of industrial intelligence (β₁ = 0.6247, p < 0.01) on high-quality green industrial development. The industrial intelligence coefficient (γ₁ = 0.2019, p < 0.01) in Column (2) is significantly positive, depicting that market-based environmental regulations and industrial intelligence are positively related to each other. The coefficient of MER (φ₂ = 0.0431, p < 0.05) is positive at a 5% level of significance; however, the coefficient of industrial intelligence (φ₁ = 0.5521, p > 0.1) in Column (3) is declined and insignificant indicating the full mediating influence of market-based environmental regulation between industrial intelligence and high-quality green industrial development. It demonstrates that stringent market-based environmental regulations encourage investing in technological advancement and equipment upgrades to encourage green industrial development. Our findings are consistent with Feng and Chen’s [48] study conclusion, which indicated that sewage charges (tools of China’s market-based environmental regulation) would raise manufacturing costs and lower company profitability. As a result, businesses employ advanced technologies to enhance or build cleaner production processes with less pollutants and low energy consumption.

The estimation results for mediating effect of administrative environmental regulations are shown in Columns 4 and 5. The combined coefficient between industrial intelligence and administrative environmental regulation is (0.3130) positively significant, but the coefficient between administrative environmental regulation and high-quality green industrial development (0.0552) does not pass the significance test. This indicates that industrial intelligence is incompatible with promoting high-quality green industrial development through administrative environmental regulation. The mechanism behind this may be that industries cannot use new advanced technical equipment in manufacturing and other operations when government departments and environmental protection agencies strictly enforce environmental rules, regulations, and policies. Next, regarding the mediating influence of public-participation environmental regulation, the relationship between industrial intelligence and PER (0.2720) is significantly positive. PER also exerts a significantly positive influence on high-quality green industrial development, as shown in Column (7) with a coefficient (0.0463). Moreover, the coefficient of Ghdi (0.5742) in Column (7) is significantly positive but less than the coefficient of Ghdi (0.6247) in Column (1), which shows that public-participation environmental regulation exerts a partial mediating influence. Explanation illustrates that improving public access to environmental information, increasing public knowledge of environmental issues, and popularizing environmental education are all advantageous in using technologies and transforming green industrial development [47]. Considering the analysis above, hypothesis 4 is partly accepted.

Table 7. Mediating impact of carbon intensity.

Variables	(8) lnGhd	(9) lnCI	(10) lnGhd
lnII	0.6247 *** (6.2610)	−0.0547 *** (−2.4120)	0.4309 ** (4.1263)
lnCI			0.2105 *** (2.4437)
Control v	yes	yes	yes
Fixed effect	yes	yes	yes
Constant	3.5517 ***	−2.0344 ***	3.2566 ***
Observation	2025	2025	2025
R2	0.831	0.7844	0.8520

*, **, and *** show the significance test at 10%, 5%, and 1%; t-values are shown in brackets.

illustrates the mediating effect of carbon intensity. The regression coefficient of industrial intelligence (β1 = 0.6247, p < 0.01) in Column (8) is significantly positive; however, the coefficient of lnII (γ1 = −0.0547, p < 0.01) is negative at a 1% level of significance which confirms that the level of carbon intensity is inhibited by industrial intelligence. The impact coefficient of corban intensity (φ_{2 =} 0.2105, p < 0.01) on high-quality green industrial development in Column (10) is positive at a 1% significance level. However, the coefficient of industrial intelligence (φ_{1 =} 0.4309, p < 0.05) in Column (10) is also significant at a 5% significance level, indicating the partial mediating influence of corban intensity between industrial intelligence and high-quality green industrial development verifying hypothesis 5. It implies that industries highly focus on using advanced intelligent technologies to reduce carbon intensity which facilitates efficient operational data collection, automatic control, and energy conservation [54], ultimately driving high-quality green industrial development.

4.4. Robustness Test

We ran a different robustness test to verify the accuracy of the findings.

Table 8. Robustness results.

Variables	(11) lnGhd	(12) lnGhd r = 0.5	(13) lnGhd r = 0.8	(14) lnGhd SDM (0–1 Matrix)	(15) lnGhd SDM (Nested Matrix)
lnII	0.0735 *** (0.0441)	0.0541 ** (0.2124)	0.0342 ** (0.1011)	0.5291 *** (2.002)	0.2411 *** (2.191)
constant	4.0331 **	3.0542 **	5.0361	1.4281 *	1.2991 **
Control variables	yes	yes	yes	yes	yes
rho				0.3131 *** (4.112)	0.2911 *** (3.362)
Direct effect				0.5289 *** (2.201)	0.2432 *** (4.253)
Indirect effect				0.0421 **(5.021)	0.0212 *** (4.261)
Total effect				0.571 **(7.298)	0.2644 *** (5.236)
observations	2025	2025	2025	2025	2025

*, **, and *** show the significance test at 10%, 5%, and 1%; t-values are shown in brackets.

presents the results. First, for a substitute explained variable, the study adopted the DDF-ML approach to quantify high-quality green industrial development [55]. The result for the alternative explained variable is shown in Column (11). Second, to look at the marginal effect, the quantile regression model is used (see Column 12 and Column 13). Additionally, the estimation outcome of a spatial econometric model may be impacted by various spatial weight matrices. To lessen the impact of matrix selection, the nested matrix and binary geographic weight matrix (0–1 matrix) were employed (see Column 14 and Column 15). The robustness checks’ findings in Table 8 show that the coefficients still match the signs of the initial regression coefficients, and industrial intelligence remains critical in promoting high-quality green industrial development.

5. Conclusions and Policy Implications

5.1. Conclusions

The study aims to examine the influencing processes of industrial intelligence driving high-quality green industrial development from a multifaceted approach, using data from 31 Chinese provinces. Here, we outline some of the study’s potential contributions: First, industrial intelligence acts as a crucial component in high-quality green industrial development, and using intelligent technologies favorably promotes the growth of green industries. Second, we also highlight the industrial intelligence’s spatial spillover effect. It is noteworthy that, through spatial spillover effects, industrial intelligence can effectively encourage high-quality green industrial development both inside the region and in nearby regions. Third, focusing on regional variations reveals the considerable impact of industrial intelligence on high-quality green industrial development in central and eastern regions, while it is resisted in the western region. Fourth, we investigate the transmission mechanisms further and discover that industrial intelligence fails to drive high-quality green industrial development via administrative environmental regulation. However, industrial intelligence can indirectly enhance high-quality green industrial development with the mediating effect of market-based environmental regulation and public-participation environmental regulation. We further discovered that carbon intensity partially mediates the relationship between industrial intelligence and high-quality green industrial development.

5.2. Policy Recommendation

The following are some crucial policy implications of the findings: First, policymakers must work to hasten technological improvements and the expansion of the artificial intelligence sector in crucial fields such as blockchain and remotely operated systems, which are expected to generate industry-wide advantages, in order to establish a strong basis for industrial intelligence. Intelligent technology and algorithms must be encouraged and regulated across a wide range of industrial disciplines, particularly in pollution prevention and energy conservation, in order to maximize the influence of industrial intelligence on green sustainable development. Second, when implementing environmental regulatory laws and mechanisms, the government should consider how well each sort of regulation will work in various geographic areas. Overall, there should be a steady decrease in the usage of administrative environmental regulating tools and policies across the country. Moreover, the government should impose environmental taxes to deter actions that potentially increase CO₂ emissions. Investors should build environmentally friendly projects to avoid carbon emissions and drive high-quality green industrial development. Third, based on the regional growth patterns and economic conditions, differentiated and customized adaptation policies should be devised. Because the western area of China has a stronger need to improve industrial intelligence than the eastern or central regions, the central government should assist the growth of intelligent enterprises in western China with funding and high-tech expertise.