1. Introduction

In response to the increasingly severe effects of global warming, many countries have agreed to promote the green transformation of their economies and societies. Reducing carbon emissions is one of the important goals of green transformation in all countries. Since the Paris Agreement, many countries have made varying degrees of effort to reduce their energy and carbon intensity per unit of GDP. However, the global warming situation and the task of reducing emissions are still significant. According to the report Climate Change 2022: Climate Change Mitigation by the IPCC Working Group III, the average annual greenhouse gas emissions in the period of 2010-2019 were higher than those in any previous decade, and all sectors need to reduce emissions immediately in order to slow down global warming. As the industrial sector produces about one-quarter of global emissions, promoting reductions in industrial emissions is crucial to fighting against global warming. China has been an active practitioner in promoting the mitigation of global warming. For example, during the 13th Five-Year Plan period, China's industrial sector achieved significant reductions in its emissions, with a cumulative reduction of 18% in CO₂ emissions per unit of industrial added value for enterprises above the scale. However, it is undeniable that behind these improvements, there is still a serious imbalance between the industrial structure and the energy structure. On the one hand, the development of a heavy industrial structure and a coal-based energy structure directly affects the speed and effectiveness of achieving the carbon peak and carbon neutrality goals; on the other hand, the potential path dependence restricts the pace of industrial technological innovation that promotes green transformation, which is not conducive to sustainable industrial development. 

With the in-depth development of the new scientific and technological revolution and the industrial transformation represented by artificial intelligence (AI), the evolution of industrial intelligence (IIG) has accelerated worldwide, leading to the transformation of manufacturing methods, such as intelligent manufacturing ^[1][9], which could have a revolutionary influence on traditional industrial development. Looking at data from industrial robot installations around the world, it is obvious that the wave of IIG is unstoppable. However, due to the potential differences in economic mechanisms between IIG and the environment, its impact on PCCR is considered to be at two different levels. The existing literature shows that there has been little research on the synergistic effects of IIG on PCCR, and even less on the mechanisms of its causal nexus. To supplement the gap in knowledge, rwesearchers set the following main research objectives:

1. To analyse the temporal and spatial evolution of carbon emissions and environmental pollution in various provinces of China.

2. To evaluate whether IIG has synergistic effects on PCCR, i.e., to discuss the economic relationship between IIG and PCCR.

3. To identify the two mechanisms of green technological innovation and energy efficiency, and to expand the study of Objective 2.

To achieve these objectives,  rwesearchers selected industrial robots as proxies for IIG to evaluate the synergistic effects of IIG on PCCR and reveal the potential mechanisms of its causal nexus based on panel data from 30 provinces in China from 2006 to 2020. The marginal contributions of this entrypaper are as follows: firstly, in contrast to the existing literature, in which studies have explored the effects of IIG from the perspective of pollution reduction or carbon reduction,  researcherswe addressed PCCR within a unified framework to focus on the synergistic effects of IIG on PCCR and examine the heterogeneity of these effects from multiple perspectives, which could further enrich the literature on the economic and environmental effects of IIG; secondly, the mechanisms of the synergistic effects of IIG on PCCR are multidimensional, and  researcherswe identified the transmission channels of green technological innovation and energy efficiency, which could enrich the literature on the synergistic effects of IIG on PCCR. This  entrypaper provides new evidence for the synergistic effects of IIG on PCCR.

2. ChLiterasing Greenture Review

The impacts that new technologies have on the environment have long been of great interest to scholars. Along with the rapid development of AI, many studies have explored its impact on the environment from different perspectives. This papentryr is closely related to two main strands of the literature: the effect of IIG on carbon emissions and the effect of IIG on pollution control.

2.1. Studies on the carbon reduction effect of IIG

Regarding the impact of IIG on carbon emissions, a study by Chui et al. found that electricity suppliers for smart grids could reduce the mismatch between electricity supply and demand by analysing the electricity consumption profile and consumption characteristics of each meter, allowing them to respond to demand more accurately. This also allowed consumers to view their electricity consumption in a timely manner using smart grid terminals and reasonably manage their use of high-power-consumption appliances. This could reduce energy waste and improve energy efficiency. Zhang and Xuan also found that connections and interactions between intelligence and applied technology departments could promote the rational allocation of energy, improve energy efficiency, reduce energy waste, and promote energy conservation and emissions reduction.

Jadoon et al. took cement plants as their research object and found that carbon-emission-monitoring algorithms that were based on AI could improve the efficiency of carbon capture at cement plants. They also established that intelligent cement plants could optimise their energy use and contribute to reductions in carbon emissions. Sankaran also pointed out that AI, as a new tool, could promote the transformation of industrial carbon emissions into green fuels and promote the development of circular economies while reducing carbon emissions. In terms of empirical studies, Liu et al. used data from China's industrial sector in 2005-2016 and found that IIG significantly reduced carbon emissions. They also discovered that this effect was heterogeneous within the industry. 

However, some studies have presented different opinions. Paryanto et al., for instance, suggested that the rapid increase in the use of industrial robots could increase energy consumption and carbon emissions, as industrial robots are the basis for automation and intelligence, but the manufacturing of industrial robots also consumes energy. Dhar also pointed out that IIG could have a two-sided impact on climate change: on the one hand, IIG could promote the building of infrastructure for reducing emissions, the use of climate change modelling and forecasting, and the control of carbon trading market prices, which could encourage industrial enterprises to reduce their carbon emissions; however, on the other hand, industrial intelligent equipment could also consume large amounts of energy, which would be converted into carbon emissions and huge electricity costs. The different impacts of IIG on reducing carbon emissions could also stem from the diffusion process and the degree of technological diffusion.

Some other scholars have discussed the mechanisms of the effect of IIG on carbon emissions. They have suggested that IIG could reduce carbon emissions by reducing energy consumption and improving energy efficiency, improving green total-factor productivity and energy intensity, and optimising industrial structures and promoting green technological innovation, among other mechanisms.

2.2. Studies on the pollution control effect of IIG

Regarding the impact of IIG on air pollution control, AI has already been used in the field of environmental monitoring for a long time, and its role in urban air pollution and environmental monitoring has been confirmed by numerous studies. AI technology has not only been used as a tool to monitor ecological changes in the environment but has also been widely applied in the industrial sector. For example, intelligent devices have been used to monitor and evaluate pollution control in real time during production processes so as to improve the operational efficiency of pollution control devices and technologies . As many intelligent manufacturing systems are highly integrated throughout supply chains, AI also has an impact on the green production behaviour of supply chain partners and industrial enterprises by promoting green technological innovation and industrial structure upgrades. Studies have typically discussed the impact of IIG on pollution emissions based on the use of industrial robots. 

2.3. Literature Gap

It can be seen from the above literature review that scholars have focused on the economic and environmental effects of IIG at two levels: pollution reduction and carbon reduction. Although different results have been reported, they all provide useful references for this study. However, regardless of opinion, the existing literature has not considered PCCR within a unified framework. As mentioned above, considering the relationship between carbon emissions and air pollutants and the call for policies to promote the cooperative governance of PCCR, it would be meaningful to study the synergistic effects of IIG on PCCR. How does IIG have synergistic effects on PCCR? What are the potential economic mechanisms? What are the possible factors that affect the synergistic effects? In this study, PCCR was examined within a unified framework, and the synergistic effects of IIG on PCCR were evaluated, which could help to further deepen IIG research and provide references to help to solve the contradiction between economic development and environmental protection.

3. Theoretical Analysis

3.1. The synergistic effects of IIG on PCCR

IIG can be regarded as the integration of AI, the industrial Internet of things (IoT), and other intelligent production technologies. Typical examples are intelligent manufacturing and intelligent factories. Therefore, the key to IIG lies in the development of intelligent technology itself, as well as the application of intelligent technology in the industrial sector and the integration of both. AI has obvious "green" characteristics in that it promotes technological innovation, equipment upgrades, and improvements in production efficiency among enterprises, which could improve the green productivity of the industrial sector. The ways in which IIG promotes PCCR are as follows:

Firstly, IIG can help to improve the prediction ability and measurement accuracy of industrial waste gas concentrations, SO₂ concentrations, temperature, humidity, and particulate matter content, as well as improving the pertinence of energy savings, consumption reduction, and pollution control. By relying on AI and big data analysis technology (BDA), smart factories can predict the energy and carbon intensity of any production process, adjust and optimise production lines, reduce ineffective and redundant operational links and, thus, reduce energy consumption. For example, by using sensors, identification analysis, and other technologies, data on electricity, heat, raw materials, and waste can quickly be collected. Through the analysis of these data, the main sources of CO₂ and air pollutants can be quickly located to form targeted solutions for carbon emissions reduction and pollution control.

Secondly, with support from big data, intelligent devices can have the function of self-learning. The historical energy consumption data for each machine can be recorded in real time, automatically analysed, and then transmitted to engineers for deep learning and continuous optimisation. Through the heuristic growth of human and machine intelligence, the functions of intelligent devices will continue to develop, thereby increasing productivity, reducing energy intensity per unit output, and improving green production efficiency.

Thirdly, enterprises can learn from one another through the knowledge spillover effect. Improvements in the intelligence of the whole industry could also lead to improvements in product quality, which would provide higher-quality raw materials and intermediate products for downstream enterprises, thereby incentivising downstream enterprises to optimise their energy allocation structures and helping to improve the energy structure ^[2][12]. These improvements in the energy structure could help to reduce carbon emissions. In addition, reducing the use of fossil fuels and increasing the use of cleaner energy sources could also reduce the emissions of other air pollutants, such as SO₂. Moreover, IIG could integrate intelligent technologies and cleaner production technologies within enterprises and industries through the industrial chain, accelerate the intelligent greening process of enterprises and, therefore, promote synchronous reductions in carbon and air pollutant emissions.

IIG is the direction of industrial development in all countries in the era of Industry 4.0, and it could also have an impact on national environmental policies at the macro level. By using intelligent tools, such as Blockchain, governments can supervise regional energy distribution and optimise the institutional mechanisms of energy trading. The in-depth application of AI in environmental monitoring, pollution control, and digital government could help governments to improve the emissions behaviour of enterprises; make up for the inadequacies of existing environmental supervision policies caused by information asymmetry; reduce energy waste, environmental pollution, and overcapacity; and promote reductions in carbon and air pollutant emissions.

3.2. Analysis of the mechanisms of IIG on PCCR

IIG not only has direct synergistic effects on PCCR but also exerts those synergistic effects through the channel of green technological innovation.

Firstly, IIG can greatly reduce the costs of green technological innovation for enterprises. Due to the path dependence of technological innovation, enterprises tend to choose technologies that can reduce costs, so technological innovation becomes limited by the existing low-efficiency technologies ^[3][36]. These technologies struggle to meet the needs of today's green production. For industrial enterprises, green technological innovation means additional costs on the basis of their existing production costs. The development of intelligent technologies, such as digital twins, could allow innovation processes to no longer be limited to physical space. The R&D, testing, and other processes of green product performance are completed in virtual space, which can greatly improve the fault tolerance rate of green technological innovation in industrial enterprises, while also reducing resource consumption in innovation processes. IIG could also reduce the labour costs for R&D. In the digital age, industrial enterprises no longer need long-term employment contracts with employees. With the help of crowdsourcing models (for example, open-source communities), it is possible to break through the barrier of geographic space and obtain immediate, professional, and diverse green solutions from home and abroad, which could significantly reduce green R&D costs. Enterprises can also obtain ideas and demand for green products from consumers, make innovation more targeted, and reduce the costs caused by new products that are not meeting market demand. When the costs of intelligence-driven innovation are reduced and new products are favoured by consumers, the motivation of enterprises to participate in green technological innovation increases.

IIG also has a substitution effect on low-skilled workers and can promote enterprises to improve their human resources, thereby providing direct human capital and knowledge sources for green technological innovation. In addition to acquiring external knowledge via open-source communication, among other methods, data can also be connected to all aspects of enterprise production with the support of intelligent devices .For instance, every machine can share, transmit, and store data in real time. A large amount of data can provide different information for enterprises. When this information is analysed and interpreted, it becomes a source of knowledge for green technological innovation.

As specialisation becomes more and more common in the era of Industry 4.0, smart factories could also innovate collaboratively with supply chain partners through business models, such as platforms. Smart factories could cooperate with supply chain partners to develop green technological innovation so as to further share its costs and risks, make up for the deficiencies of independent innovation, and improve its efficiency. Studies have shown that green technological innovation can not only directly reduce net emissions  but also reduce the costs of pollution control by making use of the innovation compensation effect. Green technological innovation also has a contact effect on supply chains, whereby green technological innovation in one sector is transmitted to other enterprises and sectors through supply chains, promoting the green transformation of whole supply chains. Green technological innovation could also promote the transformation and upgrading of industrial structures, promote the development of clean industries, and eliminate backwards production capacity, thereby forcing enterprises to optimise their energy consumption structures and reduce their carbon and air pollutant emissions.

Finally, green technological innovation in the energy sector could accelerate the development of photovoltaics, wind power, and renewable energy, effectively promote the development of a new energy sector, and be conducive to the transformation of energy consumption structures to green, low-carbon, and clean energy, which could directly reduce carbon and other air pollutant emissions.

AI is a typical technological innovation in the era of Industry 4.0 ^[4][46], which can directly encourage enterprises to optimise and upgrade their existing technologies, equipment, and processes and improve their production efficiency, thereby improving green total-factor productivity. By relying on modern information technologies, such as IoT and ERP systems, production, management, services, and other production links are not separate. Enterprises only need a small number of skilled workers to control whole production processes in real time, promoting the rational allocation of energy and other factors. Improvements in resource allocation efficiency are an important antecedent for improvements in green technological efficiency, which is beneficial to industrial green productivity. Green productivity can directly improve energy efficiency which, in turn, contributes to reductions in carbon and air pollutant emissions.

In addition to some simple automation equipment, traditional production lines are mainly operated manually by workers. There are many problems with this model, such as safety, inaccurate control, and material waste. The substitution effects of intelligent equipment, such as industrial robots, improve the accuracy of production processes, reducing the inevitable resource waste and losses caused by manual operation, and improving energy efficiency ^[5][3]. Some intelligent devices can also realise completely unmanned operation, thereby improving production and energy efficiency.

In addition to intracompany effects, the impact of IIG on energy efficiency can also be seen at the industry level. With the support of BDA, energy demand and use data can be spread quickly among enterprises, thereby promoting the rational allocation of energy within industries. As IIG expands, it could promote the formation of intelligent industrial clusters, thereby releasing the agglomeration effects of specialisation and diversification. In these intelligent industrial clusters, information regarding energy prices, transaction volume, and inventory could flow freely, and the energy efficiency could be improved. Specialised agglomeration could also promote the integration of knowledge in different enterprises, thereby helping to improve regional energy efficiency. Furthermore, improving energy efficiency could directly reduce the emission intensity and levels of carbon and air pollutants, bringing about the synergistic effect of reducing pollution and carbon emissions.

4. Conclusion and Policy implications

4.1. Conclusions

IIG is vital for addressing climate change, promoting reductions in urban carbon emissions, and reducing pollutant emissions by replacing the demand for emission-intensive products with dematerialisation and optimising resource management and decision-making processes through system integration. This  papentryr used provincial level data in China from 2006 to 2020 to test the effect and mechanisms of IIG on PCCR through empirical analysis. The conclusions are as follows:

(1) The spatial distribution characteristics of the quality of the ecological protection environment among provinces in China are prominent, showing a decreasing spatial distribution pattern from east to west;

(2) IIG could significantly reduce concentrations of CO₂ and PM_2.5 in cities, which has significant synergistic effects on PCCR;

(3) IIG could reduce pollution and carbon emissions through mechanisms of promoting green technological innovation and improving energy efficiency.

The existing literature has suggested that the expansion of production scales caused by IIG could have an adverse impact on the environment; however, theour study found that when digital technologies and the corresponding industrial intelligent tools were used to promote coordinated reductions in pollution and carbon emissions, there was a continuous promotion effect on improvements to the environment. On the one hand, it would be beneficial to reduce the costs of green technological innovation; on the other hand, new achievements in IIG could stimulate green technological innovation and promote the development of green products. Furthermore, these changes could cause the industrial sector to upgrade its industrial structure, eliminate backwards production capacity, and promote the transition towards green economies and societies.

4.2. Policy implications

Given the need for green and low carbon development, it is urgently needed to deeply integrate process industry with big data and machine learning in the future and to construct and describe the data resources related to carbon emission and carbon neutralization technologies of industrial industries. Based on these findings, researchwers put forward the following policy recommendations:

Firstly, the results of this study showed that industrial robots could be used to reduce pollution and carbon, so it is necessary to focus on developing the robotics industry and speeding up the creation of core technologies. TWe should speed up the development of essential standard IIG technologies should be speeded up, such as chip technology, sensing technology, and information technology, and carry out technical research on core components and major landmark products. Administrative departments should encourage market players to accelerate the development of market-oriented high-end robot products through financial subsidies, tax reductions and fee reductions, and the promotion of domestic robots in order to achieve technological breakthroughs.

Secondly, an energy IoT should be built, and the intelligent transformation of power systems should be promoted. Accelerating the intelligent transformation and efficient operation of energy systems could help IIG to reduce carbon emissions and protect the environment. Relevant departments should use big data, Blockchain, and 5G technologies to promote the development of intelligent and digital national power grid systems, focus on the efficient utilisation of traditional fossil-fuel-based energy, improve the development and consumption levels of renewable energy, and accelerate the construction of new power systems that mainly use clean energies. Based on the gradual establishment and improvement of software platforms and databases, industrial production's pollution reduction and carbon reduction research will minimize the total cost. With the value-added cycle of water, material, and energy as the core, information big data integration and artificial intelligence decision-making of pollutants, water, and energy at different levels can be realized. Make full use of information technology to break down the barriers between industries and fields, and help the traditional single industry to reduce pollution and carbon, and the upstream and downstream industrial chains to coordinate pollution and carbon reduction and digital pollution reduction and carbon reduction.

Thirdly, the development of green and low-carbon core technologies should be strengthened, and enterprises should be encouraged to carry out green R&D innovation. Financial institutions should actively launch diverse digital and green financial products. At the same time, administrative departments should formulate credit policy incentives and increase financial capital and support for enterprises that are involved with green technological research and development—especially those involved with carbon sequestration, carbon capture, and clean energy. Governmental departments should support the popularisation and application of green technologies in industries with high energy consumption rates and high pollutant emissions in a multilevel, multidimensional, and all-encompassing way, thereby giving full play to new knowledge regarding how to reduce pollution and carbon emissions.