Abstract：This paper focuses on how to use AI artificial intelligence to resist air pollution， and analyzes the successful application of AI in environmental engineering. At the end of this paper， the difficulties of artificial intelligence application are summarized and the bright prospect of AI in China is outlined.
Key words：AI; air pollution; environmental engineering
1 The aids of AI in improving air quality
IBM has successfully tested a brand new method with the involvement of Artificial Intelligence and big data， to alleviate choking air pollution in Beijing（China）， where the densely-populated regions are surrounded by coal-powered factories releasing toxic particulates. The air quality， however， could vary depending on various factors， e.g. traffic congestion， industrial activities and the implementation of legislation associated with environmental protection.
A giant computer system capable of learning to predict the severity of air pollution in different parts of the city several days in advance， has been invented. By means of combining and analyzing large quantities of data from several different models―an extreme computational challenge， the system could ultimately pinpoint the precise sources of air pollution. Thus， it can present specific recommendations on how to enhance the air quality reasonably to an acceptable level―for instance， by allocating tradable pollution permits or implementing taxation， and restricting the number of vehicles on road.
With reference to the intelligent system that predicts the air quality in order to provide resolutions in advance， artificial intelligence has already contributed to the improvement of air quality to a great extent.
It is generally recognized that artificial intelligence is a data-driven decision-making mechanism. Using real-time data and all kinds of information， as well as comprehensive deployment and application analysis， it ultimately achieves automatic intelligence and maximizes the benefits to the society. From driverless cars to visual reality， from mass manufacturing to online education to supporting doctors with minimally invasive surgery and logistics distribution， and so on， the powerful “deep learning” and “fast processing”capabilities behind them are indispensable for their accomplishments in various domains shown like picture 1. Consequently， how precisely can artificial intelligence be applied in environmental monitoring， the surveillance of air quality in particular？ Picture1 AI simulation diagram
Currently， the air quality monitoring network in a few countries such as China has been preliminarily completed. Having taken advantage of artificial intelligence， standard station， satellite and meteorological data could be easily integrated as long as adequate information can be obtained， resulting in the comprehensive analysis of big data. As an illustration， with the increasingly mature technology in AI， more and more accurate decision-making mechanisms of air quality management have attracted the attention of the top professionals and scientists specializing in civil engineering， big data and artificial intelligence. Based on AI to revolutionize the approach to surveillance of the air pollution level from inside out， one commercial corporation（FAIRSENSE Ltd.）in China， which applied the technology of Internet of Things（IoT）， cloud computing and big data， has constructed a platform dedicated to the analysis of big data in the universal measuring environment（AQmap?）. Similar to the majority of pollution-predicted system， the platform integrates information from various sources including geography， atmosphere and meteorology. Not only could it visualize the display of air quality monitoring data， it could also achieve horizontal and vertical analysis of the contrast of historical data on the time dimension， and realize the detection and alarm of various air pollution on the spatial dimension. These data are collected in designated areas， and the system can vividly exhibit the real-time trend of air quality and the distribution of a variety of air pollutants （e.g. carbon dioxide， sulfur oxides， carbon monoxide， nitrogen oxides and volatile organic compounds）.
2 The difficulties associated with the application of AI in improving air quality
As I’ve previously mentioned， the computational barriers have to be overcome in order to make further progress in the prediction of air pollution. Without the appropriate algorithms and advanced technology in artificial intelligence， the air quality assessment systems developed are likely to possess low reliability， which might lead to a misleading prediction result and thus， further deteriorate the atmospheric conditions. In addition， the systematic integration of station， satellite and meteorological data collected from the monitoring networks is still a problem that appears to be difficult to address. Another difficulty faced by the authorities is how to actually implement the environmental protection via the analysis of environmental big data by improving the accuracy of air pollution prediction.
3 The prospect of AI in China
With the gradual advancement in artificial intelligence， enterprises and corporations should not only take the air pollution produced into account， but also consider the improvement of ecological environment as one of the vitally important targets. As far as I’m concerned， artificial intelligence will ultimately stimulate innovation and modernize the ecological governance system.
KNIGHT W.How Artificial Intelligence can fight air pollution in China. Retrieved from MIT technology review[EB/OL].（2018-08-31）[2018-09-10].https：//www.technologyreview.com/s/540806/how-artificial-intelligence-can-fight-air-pollution-in-china/.
仪器信息网.让人工智能（AI）成为蓝天保卫战的有力抓手[EB/OL].（2018-04-02）[2018-09-10].Retrieved from 资讯中心：https：//www.instrument.com.cn/news/20180402/243454.shtml.