INTRODUCTION:

Artificial intelligence is a field of computer technology targeted at the development and establishment of computer systems which can finish tasks efficiently that are usually done by humans — in particular, things associated with people acting intelligently. Artificial intelligence is entering into many more areas of life than ever before. The most advanced trend yet is AI chips and related applications on smartphones. However, Artificial Intelligence hails from as early as the 1950s from the Dartmouth Summer Research Project on Artificial Intelligence at Dartmouth College, USA. The birth of AI goes even further back to the work of Alan Turing, Allen Newell and Herbert A. Simon.

“Air pollution is defined as the presence of toxic chemicals or compounds (including those of biological origin) in the air, at levels that pose a health risk.” In an even broader sense, air pollution is the presence of elements in the air which are almost not present or which have lower ratio in the air. This type of pollution also causes detrimental changes to the quality of life such as damaging the ozone layer or causing global war. Worldwide, bad outdoor air caused an estimated 4.2 million premature deaths in 2016, about 90 percent of them in low- and middle-income countries (Nat Geo, 2019). Climate change and pollution are on everyone’s mind these days. Artificial Intelligence is going to be at the forefront of this battle. Machines are being developed to clean the environment and reduce the damaging effects of air pollution. Software will allow these machines to differentiate between biological organisms and pollutants. Smarter software is also fighting air pollution by advancing technology from fuel burning factories. Air pollution has become the fourth-largest threat to human health behind BP, dietary risk and smoking (WeForum, 2016). It is time that AI can be integrated effectively into entire air pollution scenario.

It is predicted that AI will be able to create knowledge and enhance management of air quality and help check air pollution. Interestingly, there is no definite limit for the level of knowledge that AI might require to make such forecasts. The scope of Artificial Intelligence is always getting broader as newer breakthroughs in its field are made.

Figure 1.: Error comparison between human performance and AI

The graph above depicts how, over years, the margin for error has changed for both machines and humans. In philosophical science, a human mind is only said to be 95% accurate at a maximum capacity. Since the beginning of mankind the human mind entails to remain constant but the AI does not. Artificial intelligence have increased their plethora of knowledge over time and advanced into a realm of reality wherein they can take decisions, suggest paths and even act. (source: AVHForum)

REVIEW OF LITERATURE:

The integration of AI-based technologies for the management and control of pollution and their recent advancements were talked about by Chan C. and Huang G. (2015) The mitigation process and fuzzy logic was discussed on the base of the paper; enabling us to get a thorough knowledge of the same. Similar concepts were brushed by Yang J. (2017) who elucidated on the future challenges of AI in environmental field along with the conjoining of fuzzy logic into the field. Chau K. (2017) through their study made a practical effort into water quality modeling and the integration of KBS (knowledge based solution). Our knowledge of the environmental quality analysis system, AI-techniques and neural networks grew after a study from Oprea M. and Iliadis L. (2018) We understood the variable important differences and modelling performance (between their employed algorithms) by developing a statistical approach to compare data distribution impacting the relevance of data. This study was bought up by Bazzi T and Zohdy M. We gained a practical knowledge when we studied a conceptual paper – discussing the 13 different areas wherein machine learning can be deployed, including energy production, CO2 removal, education, solar geo-engineering, and finance – authored by Snow J. Coming to a more implementation mindset, a study by Bradley M. 2013 discussed how technology will increase our reach directly to people, how data from different sources will allow us to drill down on local issues. Also, data analytics over time would allow people to understand impacts on their health and the latest trends and change behavior while creating opportunities for prevention. Supriya R. 2016 discussed about technology such as Internet of Things (IoT) and Artificial Intelligence (AI) which can predict air pollution in a given area and suggest appropriate actions. Microsoft’s study in 2018 showed how AI can contribute in making precise projections of changes in future air quality, such as relating satellite data to data from its own monitoring points. Microsoft disclosed they will spend $50 million in "AI for Earth" in the coming five years.

DISCUSSION:

1. Scenario of Air Pollution:

Global scenario of Air Pollution - Air pollution is a chronic scourge for developing countries like India and China. Every year, a combined populous of 3 million die prematurely in these countries from household and outdoor pollutants. Global contributors of pollution like manufacturing, vehicle exhaust and coal burning are fueling the deaths of the world’s forth leading risk factor. Air pollution is one of the main reasons for premature deaths in the world. In a WHO Study in 2018, it was calculated that out of all major global health risks, presence of outdoor air pollution as fine particulate is found to be severely hazardous than previously thought. The WHO Study ranked air pollution as one of the top 10 killers in the world, with a third of the deaths occurring in Asia. In China, in 2018 alone, particulate matter pollution was the fourth most death causing factor.

Indian Scenario of Air Pollution -There is a much larger burden of disease per person in India due to the grave condition of air pollution. Since India implemented LPG in 1990, air pollution has become the second highest risk factor for diseases. A factual description of India’s air quality level could lie in the fact that 11 out of the 12 most polluted cities in the world lie in India (Dharni A. 2018). Delhi’s pollution scenario is the country’s gloomiest. The presence of particulate matter has been detected in major metro cities and their main causes are vehicle pollution and the use of coal powered plants. Studies in India have also shown that the number of people suffering from respiratory diseases and allergies has doubled in size since the 1990’s. After major studies, it was established that long time exposure of a particulate can lead to respiratory and cardiovascular diseases. The Central Pollution Control Board was a government initiative started in 1974 to conduct investigations and research and make suggestions to the government in matters related to any type of pollution.

2. Scenario of Artificial Intelligence:

Artificial Intelligence had a chequered journey through the 20^th century. But the 21^st century, with the rise of internet, is being able to help AI realize its full potential. This potential can only be limited by the limit of AI itself which is unknown. Also, Artificial Intelligence has always shown promising output in social and existential situations.

Artificial Intelligence has already spearheaded many technological advances in various fields like: I.) AI in Business: “AI is able to support three important business needs automating business processes, gaining insight through data analysis, and engaging with customers and employees.” AI in Agriculture: “AI bots in agricultural field can harvest crops at a higher volume and faster pace than human laborers. By leveraging computer vision, farmers can monitor the weed and spray them.” II.) AI in Climate Control: “AI enables us to better manage the impacts of climate change and protect the environment”

3. Air pollution and AI:

This decade has seen an upbringing rise of the integration of AI into business values and models. Its hyperlinking connection to each detail present is only possible as AI is getting smarter and smarter – boosting human efficiency and discovery. However, as expert systems are advancing and unfolding, it’s important to understand their implication on an area beyond industries and business – the Planet.

The challenge therefore puts a rhetorical shift from an “human friendly AI” to a “Earth friendly AI”. This requires complete harnessing of knowledge base systems to create a sustainability revolution. Based on a study of 80 AI applications, PwC have documented how AI is helping changing the game in the field of air pollution:

· Weather and climate science and forecasting: Deep-learning networks and energy- intensive computing have helped forecasting the complexities of ‘real world’ systems into calculations. This has accelerated forensic discoveries and prognosis.

· Smart disaster response: Inculcating simulations and studying real time data of weather events. Benefits prolong into disaster preparation, signaling early warnings and prioritizing responses.

· Autonomous and electric vehicles: Reduction of greenhouse gas emittance and dealing with ecological driving thereby setting a transition for the benefit of generations to come.

· Optimizing energy sources: Substantiating the demand and supply of energy sources, AI has helped prolonging the weather conditions by integrating the reliability of such power source for a global incentive.

4. Stated case based reasoning:

To comprehend and adjust to the present situation of air pollution better, we should gain from our outside/external environment and acknowledge the great which is being done by some driving IT organizations. Here are some current case tests/samples which are being finished by them, from which we can learn different procedures and learnings about how to actualize and coordinate AI with social issues and ration and move towards a feasible and sustainable objective in future.

1. IBM aids in battling Beijing's air pollution emergency:

Beijing, China had two peculiar cases of reaching critical levels of air quality in 2015. A study by Berkeley Earth induced that the air levels present in Beijing were comparable to smoking 30 cigarettes a day. IBM, then with their Green Horizon initiative ought to integrate IoT and AI in order to predict AIQ levels and also lower the pollutants. Connected sensors were affixed all over China’s capital which would identify all factors contributing to Beijing’s climatic deterioration. By 2017, the air nature of the city radically improved – a per cubic meter of 2.5 concentration in air quality levels. IBM’s initiative was termed successful as their data study not only helped in preparing for future times to come but also reduce the emission of harmful smog particles by 25% in 2017.

2. Romania’s fight against Radon ailed by Dositracker:

Radon is a radioactive gas that could cause lung cancer and is released most commonly by dwellings and building. Whilst Romania were furthermore concerned about the emission of Radon, a Bucharest based startup – Dositracker were able to tackle this problem by their pragmatic project called the RadonAir. These data collecting tools were added to all construction sites and mining operations as a way to efficiently use resources with minimal emission of radon in the city.

3. UAE opens AI labs to forecast AQI in advance:

An initiative of Ministry of Climate Change and Environment, a satellite-monitoring based Lab was launched which would allow authorities to get real time data and also identify the origin of air pollutant. A mere tool used for anticipation, the system would predict dust storms and plot down the sources of pollutants to be better prepared for a maximum future time interval of 3 days. Moreover, an accessible app was built was residents to view the current and future scope of air quality. The lab studies conducted located the sources of C02 in transport, industrial undertakings and even service sectors, which overall helped the government in establishing the correct policies and regulations on the most prominent segment.

Analysis:

i. Visual analysis:

The fig (2) shows and depicts the level of pollution varying across all the states in the nation, it is measure of the aggregate air pollution in existing eco system, which is pictographically shown in the above picture. Hence it delineates the degree of contamination differing over each one of the states in the country, it is proportion of the total air contamination in existing eco framework, which is descriptively shown in the above picture. (Data source: Kaggle)

Another vital visualization to refer to is the heat density map, which shows the intensity of pollution across the country and the region which is highly contaminated is represented with a dark red color on the map. This is an essential representation to allude to pattern, which patterns the contamination over a nation and the locale.

Figure 4: Heat map depicting level of pollution per State

ii. Discussion:

1. Scrutiny of the topic:

Many authors derived various conclusions which all align in the same direction that if at all, the major tech giants, central government and individuals at their part try to incorporate the significant role of artificial intelligence into various industries, then it would not only aid in curbing the air pollution levels all across the country but also act as an early remedy for various other natural and man-made events which may prove deadly for the mankind. Also, we agree the implementation phase of the same may be very burdensome because of lack of proper infrastructure and prevailing loop holes in our nation, so if someone questions the feasibility of the implementation, it wouldn't be wrong because we genuinely lack major essentials which may be needed to execute such ventures across the nation. But as a duty of a good citizen, we need to take such initiates.

2. Barriers:

The paper induces what are the obstacles present in our nation’s perimeter as to how we’d apply artificial intelligence into the existing climate scenario. From cultural norms to lack of a communist approach embedded in the same, the following points are presented which are potential barriers for India in its track to apply AI: -

Lack of enabling data ecosystems: Simply put, the current data situation of India is not very promising for Artificial Intelligence to grow as a portfolio in various industries. ‘Enabling Data’ refers to the required volume of data that AI tools will utilize to efficiently produce an output. This output, however, can be quantitative or qualitative depending on the task performed.

Low intensity of AI research: India is a developing country which has huge potential for Artificial Intelligence in every major sector. But, the level of core research being conducted for fundamental research is quite low and has also slowed the innovations in the field. Also, most companies are unsuccessful in completing transformation of core research into applications for the market.

Inadequate availability of AI expertise, manpower and skilling opportunities: Artificial Intelligence is a complex concept which requires expertise and manpower to master. However, India has not yet reached that level of skilling opportunities that it requires to compete on a global scale.

High resource cost and low awareness for adopting AI in business processes: Cost of implementation of Artificially Intelligent business processes has been increasing in the last decade in India. This is due to the lowered awareness of these functions arising out of lack of investment in the same.

Cultural barriers: Human beings are famous for resisting change. We are creatures of habit and once we develop a system of conducting a mission that seems to effectively and efficiently get the job done, we prefer it. It often takes some encouragement before we witness the total gains Artificial Intelligence will bring. Also we will then understand the worth of the disruption and expense that will inevitably be caused by changing procedures or adopting new processes.

3. Application:

Studying the underlying practicality of the subject is important. Our paper incites the course of functionality of artificial intelligence in the scope of the nation’s existing problem with any pollution for that matter. They are discussed as follows:

· Real time integrated, adaptive urban management. The metropolitan cities account for more than 52% of the total air pollutants present in India. It’s time for big authorities to take an integrated computation approach which would stimulate real time data to provide efficient courses of action.

· Fuzzy control implemented into agricultural process. Agriculture is the second biggest sector contributing to the India’s GDP; the emittance of heat can be reduced given fuzzy control is enabled at horticulture plateaus.

· Sensors integrated with machine learning and network technology. Such automations have benefits of early identifying gas leaks, contaminations, and epidemics. An Australian local body professed that the lack of sensory devices had enlarged minor bush fires which could have been earlier noticed.

· Data backed resilient planning. A political figure in Kerela confessed that lack of resilient plans made matter matters uncontrollable and worse. Any contingency ideas backed by AI encompasses for social, environmental and economic problems into goals and strategies.

· We all are aware of the risks associated with AI – in terms of privacy, security, economic and social factors. The government of India needs to ensure a transparent cycle of same. Cohesiveness with policymakers and industrialists is required to relieve the issues of AI and optimally utilize it.

CONCLUSION:

Undoubtedly, most developed nations have intended to apply AI tools and techniques into their environmental policies today. Our case based study on Beijing showed how integration of AI helped them decrease the pollutant concentration in their ozone. China too has taken huge steps to improve their air quality since the beginning of 2017 – with the help of adaptive machine learning.

There are a remarkable number of cases with AI being implemented to improve a demo graph’s ecosystem. India, with currently one of most degraded air quality faced the eventuality. Its capital region, Delhi recorded an AIQ of 499 in November 2019, which for record falls into the ‘severe’ category. With daily emittance of greenhouse cases and carelessness on our parts; its important to raise the question of What are we going to do about? Perhaps, with rising interconnectivity with AI (IoT) and machine learning, the realms of possibility are interminable. Our study put the light on India’s problem with sustainability in the field of pollution, and for all right reasons the benefits bought by Artificial Intelligence as a branch should not be ignored.

REFERENCES:

1. Bradley, M. (2019, June 5). #BeatAirPollution - How AI and IoT could help people combat air pollution issues. Retrieved from https://www.capgemini.com/2019/06/beatairpollution-how-ai-and-iot-could-help-people-combat-air-pollution-issues/

2. Chau, K. (2006, August 1). A review on integration of artificial intelligence into water quality modeling. Retrieved from https://www.researchgate.net/publication/7018270_A_review_on_integration_of_artificial_intelligence_into_water_quality_modelling

3. Chan, C., and Huang, G. (2003, March 1). Artificial intelligence for management and control of pollution minimization and mitigation processes. Retrieved from https:// www.researchgate.net/publication/222670520_Artificial_intelligence_for_management_and_control_of_pollution_minimization_and_mitigation_processes

4. Hannah Ritchie and Max Roser (2020) - "Air Pollution". Published online at OurWorldInData.org. Retrieved from: 'https:// ourworldindata.org/air-pollution'

5. Mckay, Rich. (2018). Can AI help reduce Beijing’s smog? Retrieved from https://www.ibm.com/blogs/industries/can-ai-help-reduce-beijings-smog/

6. Marr, Bernard. (2018, February). The Amazing Ways We Can Use AI to Tackle Climate Change. Retrieved from https://www.forbes.com/sites/bernardmarr/2018/02/21/the-amazing-ways-we-can-use-ai-to-tackle-climate-change/#7e6b5be54a35

7. Microsoft (2019, August 20). A fresher breeze: How AI can help improve air quality. Retrieved from https://news.microsoft.com/ europe/2019/08/20/a-fresher-breeze-how-ai-can-help-improve-air-quality/

8. Oprea, M., and Iliadis, L. (2011). An Artificial Intelligence-Based Environment Quality Analysis System. Engineering Applications of Neural Networks, 499-508. doi:10.1007/978-3-642-23957-1_55

9. Poon, Linda. (2015, September). Beijing's Newest Weapon Against Smog: A Supercomputer. Retrieved from https:// www.citylab.com/life/2015/09/beijings-newest-weapon-against-smog-a-supercomputer/403306/

10. Rawat, Shubhankar. (2019, June). India Air Quality Data Analysis. Retrieved from https://towardsdatascience.com/india-air-pollution-data-analysis-bd7dbfe93841

11. Srivastava, Chavi and Singh, Shyamli and Singh, Amit. (2018). Estimation of Air Pollution in Delhi Using Machine Learning Techniques. 304-309. 10.1109/GUCON.2018.8675022.

12. Supriya, R. (2018, October 31). Delhi Air Pollution: Artificial Intelligence and IoT Can Play the Savior’s Role. Retrieved from https://www.dqindia.com/delhi-air-pollution-artificial-intelligence-iot-can-play-saviors-role/

13. Snow, J. (2018, July 18). How artificial intelligence can tackle climate change. Retrieved from https:// www.nationalgeographic.com/environment/2019/07/artificial-intelligence-climate-change/

14. Varun, P. (2019, November). India air pollution at 'unbearable levels', Delhi minister says. BBC. Retrieved from https:// www.bbc.com/news/world-asia-india-50280390

15. Ye, Z., Yang, J., Zhong, N., Tu, X., Jia, J., and Wang, J. (2020). Tackling environmental challenges in pollution controls using artificial intelligence: A review. Science of The Total Environment, 699, 134279. doi:10..016/j.scitotenv.2019.134279

16. Yetilmezsoy, K., Ozkaya, B., and Cakmakci, M. (2011). Artificial intelligence-based prediction models for environmental engineering. Neural Network World, 21(3), 193.

Received on 11.04.2020 Modified on 24.04.2020

Asian Journal of Management. 2020;11(3):285-290.

DOI: 10.5958/2321-5763.2020.00044.X